Sessions | MakeIT 2026 & JCON Slovenia 2026

Room B Wed 27 May 2026 · 09:30 - 12:30

Langchain4J CDI in Action

In 2025, we introduced LangChain4J-CDI, a tool designed for developers to integrate LangChain4J into the enterprise application, powered by JakartaEE CDI.
In this workshop, we'll showcase Langchain4J CDI, a CDI extension that leverages the power of CDI to integrate with developers' AI agents through Langchain4J AI Services.
We'll build a simple chatbot to showcase a simple AI agent.
Using LangChain4J toolbox, we'll configure a:
* ChatModel
* ChatMemory
* Memory provider (for RAG)
Using the configuration model and easily integrate it into your AI services.
We'll also use MicroProfile suites, such as Fault Tolerance and Telemetry, to record metrics of your LLM usage with ease.

Buhake Sindi

Room A Wed 27 May 2026 · 09:30 - 12:30

Fast Track Your Upgrade to Oracle AI Database 26ai

Join this intensive Hands-On Lab to build deep technical expertise in upgrading to Oracle AI Database 26ai.

* Learn how to modernize your environments migrating to Oracle Multitenant architecture
* Perform various automated tasks with AutoUpgrade for reliable, repeatable upgrades
* Explore refreshable PDB clones to streamline testing and reduce downtime.
* Run real-world workloads, compare performance, and validate behavior.

Oracle experts will provide guidance, tips, and best practices to help you plan, execute, and optimize your upgrade path with minimal risk and maximum efficiency.

Part of the workshop is doing exercises in a lab environment using Oracle LiveLabs. Be sure to bring a laptop.

Become an Upgrade hero in no time.

Daniel Overby Hansen

Room C Wed 27 May 2026 · 09:30 - 12:30

Unified Namespace v praksi: od realnočasnih podatkov do naravnega jezika

Delavnica predstavi sodoben pristop k upravljanju in uporabi realnočasnih industrijskih podatkov z uporabo Unified Namespace (UNS) kot osrednjega podatkovnega sloja.

Na poenostavljenem primeru »Acroni v malem« bomo prikazali, kako se podatki iz heterogenih virov (agregati, ERP, MES, kadrovski sistemi) zbirajo, normalizirajo in objavljajo v UNS strukturi. Podatki bodo simulirani v realnem času in sproti vizualizirani.

V zaključnem delu delavnice bomo pokazali, kako smo v UNS orodje integrirali Tools API za t. i. tool / API calling, kar omogoča poizvedbe po strukturi in podatkih UNS z uporabo naravnega jezika. Uporabnik lahko sprašuje po strukturi UNS, realnočasnih in zgodovinskih podatkih ter razpoložljivih API-jih.

Rešitve temeljijo na treh lastnih produktih: UnsDataHub, uns-archiver in uns-kit, ki skupaj tvorijo celovit ekosistem za delo z realnočasnimi industrijskimi podatki.

Aljoša Vister

Kristjan Cuznar

Room D Wed 27 May 2026 · 09:30 - 10:15

Getting started with Property Graphs - A Jumpstart for Database Developers

This session provides Oracle Database developers with a jumpstart on working with Property Graphs, focusing on the SQL/PGQ support introduced in ISO/IEC 9075-16:2023. You will gain a foundational understanding of Property Graph concepts and learn how to create graphs from existing relational data. The session covers basic graph creation, querying techniques, and practical use cases for Property Graphs. Additionally, we’ll explore how Oracle APEX enables developers to quickly build applications that offer intuitive graph interactions, empowering users to visualize and navigate complex data relationships effortlessly.

Karin Patenge

Room D Wed 27 May 2026 · 10:20 - 11:05

Property Graphs in Oracle: Your Data Has Relationships… Let’s Talk About Them

Let’s be honest: some data models are basically screaming “I’m a graph!” while we keep shoving them into relational tables like it’s 1999. Thankfully, Oracle Database includes first-class property graph support, PGQL for querying, and features that make graph modeling far less scary than its academic reputation.

In this session, we’ll demystify property graphs through relatable examples, questionable metaphors, and at least one diagram that looks like spaghetti but is technically accurate. You’ll create graph views over relational data, run PGQL to explore connections, and see how to bring those results to life in Oracle APEX — from graph visualizations to interactive drill-downs.

If you’ve ever written a join with more than six tables and thought “there must be a better way,” congratulations — you’re ready for graphs.

Richard Martens

Room D Wed 27 May 2026 · 11:10 - 11:55

Visualizing PL/SQL with PL/Scope, 26ai Property Graph and APEX Graph Plug-in

Using PL/Scope we can collect information about every identifier in your PL/SQL code compiled in the database - where is it declared and defined and where is it assigned, called or referenced. We can query that information to find out lots of interesting information about our code, but it can become convoluted hierarchical queries if we want to follow the entire call stack.

Using 26ai native Property Graph in the database, we can treat such data as a graph and query "paths" of code calls with SQL/PGQ instead of hierarchical queries. These can be visualized using the APEX Graph Visualization Plug-in.

The visualization will show the code in the database, no matter if the application front is APEX or some other client, so we can use this on any PL/SQL based application. If it's an APEX application, remember that PL/Scope does not catch code embedded in for example APEX pages - to take advantage of this technique, we should place our code in packages.

Let's try and see how much knowledge about our code we can gain in this way. Do we have unused code anywhere? Can it help impact analysis if we want to change some code? Will the visualization show us whether we have nicely modularized code or a heap of tangled spaghetti?

Kim Berg Hansen

Room D Wed 27 May 2026 · 12:00 - 12:45

What's so special about Spatial?

Oracle Spatial opens new dimensions to your data. It is a mature component in the database, but has over time gotten quite complex. This presentation for DBAs and developers will give you a gentle introduction to the basics, address typical challenges, and give you answers to questions like: How do you find your nearest bar? How do you index spatial data? Coordinate systems - how difficult can it be? Why did my house ended up in the ocean when I tried to display it on a map?
Learn how to display data on maps in APEX, so easy that even DBAs can do it.

Øyvind Isene

Room B Wed 27 May 2026 · 14:00 - 17:00

How does a Java Agent work? Building a Java agent from scratch.

Many developers have used a Java agent before, be it in your mocking frameworks, to find a memory leak, or maybe you have some sort of tracing on your backend via Opentelemetry. But what is a Java Agent, and how does it function?
Would you like to know how to manipulate run-time code and achieve things that are impossible otherwise?

In my day-to-day job, I mainly work with Java agents, and in this workshop, I will show you what an Agent can do, how they work, and how you can write on yourself.

We will set up a project with a simple Agent and use run-time code generation to automatically observe when methods are called and with what parameters. We will learn what Java classes look like when they are compiled to bytecode and how you can read them.

Marco Sussitz

Room A Wed 27 May 2026 · 14:00 - 17:00

Maintaining legacy applications using AI tooling

This session will be focusing on building a working knowledge of available AI tooling that can help developers be more productive in their day-to-day work, with a focus on existing applications and their evolution (including old applications). We'll be talking about techniques and methods to refactor and update your applications, while keeping things manageable, readable and to the level of code quality that is expected from an enterprise application.

We'll also be discussing the MCP protocol and how it can be used in order to supercharge your developer tooling even further. We'll be exploring how to use existing resources in order to e.g. obtain better information for your design, code implementation, API endpoints, etc. Then we'll also explore what it takes to build one of these servers in case you need to offer an MCP server to the rest of your colleagues.

All of these topics will be introduced through short presentations, then participants will have the option of working on illustrative examples on their own.

Cristian Schuszter

Room C Wed 27 May 2026 · 14:00 - 17:00

Analytic Functions HOL

Have you learned the syntax of analytic functions? Seen many presenters demonstrate with nice short examples how smart analytic functions are? And then went home and thought: “Well, they may be smart, but how can I use them in my real work?”

I’ve used analytic functions in thousands of SQL statements in more than 25 years developing applications. I can’t imagine coding SQL without analytic functions, as they allow me to put much functionality in SQL that otherwise might have needed procedural looping. Boss is happy as well, as the application perform much faster.

In this Hands-On-Lab I’ll showcase several of the real life use cases of analytic functions I’ve made over the years, hopefully serving as inspiration to your own work.

You'll do the lab on your own laptop on freesql.com.
The lab will contain 6 modules:

- The Basics
- TOP-N and Ratios
- Group Consecutive Data
- FIFO Picking
- FIFO Batch Picking
- Forecasting

Each module teaches a different usage of analytic functions.
Modules are beginner to intermediate level and can be taken independently - you can continue the lab at home if you run out of time.

Kim Berg Hansen

Room D Wed 27 May 2026 · 14:00 - 14:20

APEX postane graf

Ali se kot APEX razvijalec prepoznaš v naslednji situaciji?

Spremeniti moraš PL/SQL paket na bazi, vendar ne veš, kako bo ta sprememba vplivala na tvoje APEX aplikacije – ali na aplikacije tvojih sodelavcev.

V 6 minutah bom predstavil prototip aplikacije, ki to težavo naslavlja z uporabo grafa znanja in ščepca umetne inteligence.

Aplikacija samodejno zgradi mrežo povezav med APEX stranmi, procesi, PL/SQL paketi in tabelami ter jih prikaže v interaktivnem grafu znanja.
Klik na posamezen paket takoj prikaže vse APEX strani, ki ga uporabljajo, ter omogoča hiter vpogled v odvisnosti in potencialne vplive sprememb.

Rešitev je odprtokodna, delujoča in javno dostopna:

GitHub: https://github.com/rcm72/insightviewer

Predstavitev: https://rcm72.github.io/insightviewer

Projekt je nastal kot odgovor na realen problem pomanjkanja ažurne in zanesljive dokumentacije APEX aplikacij, zato potrebujem povratne informacije tistih, ki ta problem poznajo najbolje – vas.

Predstavitev je namenjena APEX razvijalcem in vodjem, ki so naveličani izgubljati čas z ročnim iskanjem odvisnosti in želijo boljši pregled nad tem, kako njihovi sistemi v resnici delujejo.

Robert Čmrlec

Room D Wed 27 May 2026 · 14:25 - 15:10

No-Code Location Intelligence with Oracle Spatial Studio

See how Oracle Spatial Studio turns raw location data into actionable insight—fast.

This session showcases an end-to-end workflow in Oracle Spatial Studio: seamlessly loading spatial datasets, geocoding addresses, performing location-based analyses, dynamically visualizing results on interactive maps, and publishing insights for use in other applications. Attendees will see how Spatial Studio streamlines geospatial projects and accelerates time to value—delivering actionable intelligence without requiring specialized GIS skills.

Renée Wikestad

Room D Wed 27 May 2026 · 15:20 - 16:05

Unlocking Insights using Spatial AI

Using machine learning with spatial algorithms and applying them to geospatial information enables systems and users to understand, reason about, and make decisions based on spatial relationships and patterns. By accounting for the effects of location, it can improve model quality and prediction accuracy. Those algorithms can be applied to analyse and interpret data derived from internal or external sources like satellite imagery, GPS, and geographic information systems (GIS), for example, to recognize objects in images and understand scenes, or to give personalized recommendations based on location information. This session gives an overview of the algorithms available with OML4Py and shows examples of how they can be used.

Renée Wikestad

Room D Wed 27 May 2026 · 16:15 - 17:00

Web GIS in Oracle APEX: Maps Without the GIS Headache

Oracle APEX is best known for rapidly building data-driven applications, but it turns out it can also handle serious Web GIS—without turning your project into a full-time cartography exercise. This presentation shows how a modern, high-performance web-based GIS can be built mostly with standard Oracle APEX components, Oracle Spatial, and open OGC services.

Using a real-world implementation, we walk through an architecture that combines Oracle APEX, Oracle Database Spatial features, MapLibre JavaScript, and GeoServer. Along the way, we tackle familiar GIS pain points: displaying thousands of geometries without performance issues, mixing internal and external WMS/WFS layers (and we hope, APEX 26.1 vector tiles), and enabling users to filter, intersect, draw, and save spatial data—all while keeping the application responsive and maintainable.

The talk focuses on practical design patterns rather than theory, including smart use of temporary tables and database views for spatial filtering, and a clear division of labor between the database, APEX, and the map client. The result is a surprisingly rich feature set: layer control, legends, spatial search, drawing tools, feature identification, data export, and automated map reports.

Alexis Zrimec

Room B Thu 28 May 2026 · 09:00 - 09:45

Is VS Code THE new Tool for APEX Developers?

Over the recent years, Visual Studio Code (VS Code) has become the most popular code editor among software developers, with its popularity spanning across different programming languages and development tasks.
Following the introduction of the Oracle Developer Tools for VS Code extension, the tool starts to be used more and more also by Oracle Database developers, thanks to its provided features
Traditionally, the APEX developers rely on the web-based APEX App Builder for application development.
Will the VS Code become suitable for full-stack APEX development, including UI design, page processing code, dynamic action and edit of static files? Will the promised APEXLang facilitate in this adoption?
This session invites APEX developers to investigate whether VS Code can effectively complement or replace traditional APEX development workflows.
Let's explore together the potential advantages and limitations of integrating VS Code into our APEX development process.

Mirela Ardelean

Room A Thu 28 May 2026 · 09:00 - 09:45

The Semantic Model: Still the Smartest Tool in the (Oracle Analytics) Room

From the very beginning, the semantic model (previously RPD) is one of the core components of Oracle Analytics. The place where various sources are modeled in such a way as to expose the business meaning of them, allowing users to fully benefit from their analytical platform. With the introduction of datasets some years ago, the modelling part has often been ignored, with users limiting themselves to simple joins and not much more.

Do you know that the semantic model allows you to turn snowflake-designed sources into simple start-schemas of measures and dimensions of attributes? Do you know that you can easily handle sources with pre-aggregated measures to make analysis a lot faster by avoiding large aggregations on the fly? Do you know that you can join together multiple sources sharing some conformed dimensions and other dimensions with no link at all with those measures seamlessly? Or, do you know that you federate various sources of different kinds and origins? Do you know how powerful logical table sources and content levels are?

If you don’t, you aren’t alone. Oracle isn’t really talking much about this piece of the product, mostly because it isn’t aligned with the buzzwording trend of self-service. The semantic model is still mostly connected to a governed analytics approach.

By the end of this session, attendees will not be data modelling gods, but will have the key elements to explore and practice the most useful techniques.

Gianni Ceresa

Room C Thu 28 May 2026 · 09:00 - 09:45

FSFO like a PRO - 19c and 26ai

Fast Start Failover is an additional configuration of the Data Guard allowing automated activation of the standby database in case of the failure.
This presentation will focus on the configuration of the FSFO at scale using Oracle 23ai database. The typical architecture and configuration will be described in details. It will cover the use of the Data Guard broker and the DBMS_DG PL/SQL package, to manage and monitor failover configurations in the different scenarios. Key aspects such as the role of the observer process, failover thresholds, and protection modes will be explained, along with practical examples demonstrating automated failover initiation and management at scale.

Marcin Przepiorowski

Room D Thu 28 May 2026 · 09:00 - 09:45

Beyond Keywords: Understanding how AI Vector Search Works

Join this talk and learn how vector search powers modern AI-driven retrieval. We'll look at what vector embeddings actually are, how they represent meaning, and why vector similarity works better than keyword matching in many scenarios. You'll see how vectors are stored, indexed, and compared inside a database, and how these pieces come together to answer queries based on intent rather than exact terms. You'll gain a clear mental model of the concepts behind semantic search so you can understand when and why to use it in real systems.

Alejandro Duarte

Room E Thu 28 May 2026 · 09:00 - 09:45

Spring Boot vs Quarkus - A Comparison

In this talk, I compare Spring Boot and Quarkus, two of the leading, modern frameworks in Java development.

I will present the core functionalities of both frameworks and demonstrate how they differ in terms of functionality and development. A key aspect of my comparison is the suitability of both frameworks about future-proofing, as well as their suitability for migrating legacy software. Using a practical example from my work, I demonstrate the implementation of an application in both frameworks to highlight the differences in practice. This presentation aims to help other Java developers gain insights into the advantages and disadvantages of Spring Boot and Quarkus.

Christian Grümme

Room F Thu 28 May 2026 · 09:00 - 09:45

Congratulations, you are agile. Your product is still late.

Agile was supposed to fix everything. Faster delivery. Better collaboration. Happier customers. That’s what the consultants promised. That’s what the slide decks still say.

And yet… your product is still late. Still not what the customer asked for. Still somehow worse than last quarter.

Companies keep burning millions on “Agile transformations” that never end. Teams drown in meetings, rituals and ticket admin. Leadership still doesn’t know what’s going on as they just use different words for the same confusion.

Scrum, Kanban, SAFe: they were meant to be tools, not religions. But somewhere along the way, organizations stopped solving problems and started worshiping the process itself.

I’ve watched teams debate backlog grooming for an hour without fixing a single user issue. I’ve seen leadership measure velocity like it’s revenue. And I’ve seen brilliant teams suffocated by “agile ceremonies” instead of being supported by a structure that helps them deliver.

If your organization is chaotic, misaligned or hiding behind processes, Agile won’t save you. Customers don’t care what framework you use. They care whether you solve their problems.

So this session shows you how to strip Agile back down to what actually matters - solving customer problems and being able to react to the market - all while having a great time building a great product. Because you do not need Scrum - you need to ship a working product.

Marin Niehues

Room B Thu 28 May 2026 · 10:10 - 10:55

How Hackers Get In: APEX Security Mistakes You Might Be Making

In this session, I will walk you through essential security measures every developer should follow when building Oracle APEX applications. We'll explore how even a small oversight or impatience can open the door for attackers to access your database. Through real-world examples and demonstrations, you'll see how common vulnerabilities can be exploited - and more importantly, how to prevent them.

To wrap up, I’ll showcase a lightweight tool designed to help you analyze the security posture and best practices of your APEX apps, making it easier to identify and fix potential weaknesses before they become threats.

Petar Simic

Room A Thu 28 May 2026 · 10:10 - 10:55

Oracle Analytics brez mitov: realni primeri, resnične koristi in kaj je novega

Predavanje je namenjeno razvijalcem in končnim uporabnikom, ki želijo iz podatkov dobiti hitre, razumljive, uporabne in pregledne odgovore.
Na praktičnih primerih bomo razbili najpogostejše mite o analitiki in pokazali, kako Oracle Analytics danes omogoča:
samostojno raziskovanje podatkov brez odvisnosti od IT,
interaktivne in prilagodljive poglede namesto statičnih poročil,
uporabo najnovejših funkcionalnosti,
hitrejše in boljše vsakodnevne odločitve.
Poseben poudarek bo na novostih Oracle Analytics in na tem, kako jih lahko uporabniki takoj vključijo v svoje delo – brez zapletenega učenja ali dodatnega razvoja.
Predavanje je zasnovano praktično, z realnimi primeri in jasnim odgovorom na vprašanje:
kaj Oracle Analytics danes dejansko ponuja končnim uporabnikom in kakšno vrednost prinaša v praksi.

Mojca Gros

Room C Thu 28 May 2026 · 10:10 - 10:55

Has my database been compromised?

Any piece of code is susceptible to malicious modifications, and PL/SQL code is not immune to it. Yes, believe me, you can have a malware running in your database at this moment disguised as a SYS package, trigger or job! In this session you will learn how to detect and clean unwanted changes in your Oracle Database core objects, highlighting security tools that can be used to track and detect such changes. Get a comprehensive understanding of how attackers explore Oracle's products and how you can use Oracle provided security tools to create an effective defense.

Rodrigo Jorge

Room D Thu 28 May 2026 · 10:10 - 10:55

Deploying a national JavaEE solution with old-fashioned VMs

How we went from a single project for one medical centre to a national healthcare solution connecting 74 hospitals and growing, using JavaEE, PostgreSQL, Keycloak and Wildfly on VMs. Managing the environment, installations and configuration with Ansible. Why did we decide to not switch to containers, what are the benefits and constraints of working with a more classic infrastructure.

Matej Zeleznik

Room E Thu 28 May 2026 · 10:10 - 10:55

Building Quarkus Extensions: A Practical Guide for Java Developers

Quarkus is often used as a framework for building fast, cloud-native Java applications, but it also provides a powerful extension model that allows you to integrate libraries, platforms and tools in a clean and idiomatic way. This session explores what are Quarkus extensions, how they work and how Java developers can start building them effectively.

We’ll begin by looking at what problems Quarkus extensions solve and when creating one is the right choice compared to other integration approaches. From there we’ll walk through the structure of an extension project and explain how build-time and runtime components collaborate to deliver Quarkus fast startup and low memory usage.

Using a practical example we’ll demonstrate how to set up an extension, wire in a third-party library, expose configuration and verify behavior with tests. Along the way we’ll cover common pitfalls, helpful tooling and design practices that make extensions easier to maintain and evolve.

By the end of this talk, you’ll have a clear understanding of the Quarkus extension model, the confidence to explore extension development on your own and a solid foundation for turning application-level integrations into reusable, Quarkus-native extensions.

Petro Velychko

Room F Thu 28 May 2026 · 10:10 - 10:55

Inside Agentic AI: Deep Dive into Modern Orchestration Frameworks

Agentic AI is reshaping how we design and structure modern software systems — from classic pipelines to dynamic networks of collaborating AI agents. But how do orchestration frameworks like LangChain, AutoGen, or Letta actually work under the hood? In this session, we’ll take a compact yet deep architectural deep dive into today’s leading Agentic-AI frameworks. We’ll dissect their internal architectures, execution flows, and role models, and uncover how they coordinate reasoning, actions, and tool usage across agents.

Beyond the surface APIs, we’ll explore real-world engineering challenges:
- How do frameworks manage state and context across multi-agent interactions?
- What are the trade-offs in error handling, scalability, and security?
- And where do their architectural limits start to show?

By the end, you’ll understand the key design principles and architectural decisions behind Agentic-AI frameworks — and be ready to decide whether to build on them or roll your own orchestration layer.

Ingo Düppe

Room B Thu 28 May 2026 · 11:05 - 11:50

Inside APEX Items: How They Work and Why It Matters

Oracle APEX items are essential components of every form page we create. But how well do we truly understand their behavior? In this beginner-friendly yet informative session, I will revisit the fundamentals and build a deeper understanding of APEX items: what they are, how they function, and how to avoid common mistakes.

This session is ideal for beginners, but even seasoned developers may discover new insights. If you've ever been puzzled by the behavior of APEX items, this talk is for you.

Lucas Hirschegger

Room A Thu 28 May 2026 · 11:05 - 11:50

AI meets Oracle Analytics

Artificial Intelligence is becoming a core capability of modern Oracle Analytics, and understanding how to use it effectively is increasingly important for Oracle professionals. In this session, we explore how AI is integrated into Oracle Analytics and how it can be applied in real-world enterprise scenarios.

The presentation covers key AI topics relevant to the Oracle ecosystem, including Oracle Analytics AI Assistant, AI Agents, and the Model Context Protocol (MCP) server. We will explain how these components work together with OAC to extend traditional analytics with natural language interaction, automation, and intelligent insights, while maintaining enterprise-grade security and governance.

The session is highly practical and demo-driven. All topics are supported by live demonstrations showing how AI features are used inside Oracle Analytics, how AI Agents can be orchestrated, and how MCP enables controlled integration between analytics, AI models, and Oracle data platforms.

This session is intended for Oracle Analytics, OCI, and database professionals who want to understand where AI fits into Oracle’s analytics strategy today, what is already available, and how to start using these capabilities in their own environments.

Žiga Vaupot

Room C Thu 28 May 2026 · 11:05 - 11:50

Upgrade Internals

Oracle AutoUpgrade has made upgrades so easy that there really is no need to know the inner workings of an Oracle Database upgrade. But we’re techies – and techies are curious and want to tear that black box apart and know about bits and pieces inside.

This session covers how an Oracle Database upgrade works and how the upgrade engine changes the data dictionary. It also explains the newer method called Replay Upgrade (which you might be using already without knowing). Further, the session describes the “lightweight” AWR that is active during an upgrade, where the normal AWR is inactive. This will help us shed some more light on what happens during an upgrade. Finally, the session shows you some advanced features in AutoUpgrade and how you can use them to tweak an upgrade.

If you’ve never upgraded an Oracle Database before you can gain valuable insights from this session. Plus, experienced “upgraders” can learn a few more tricks to put in their toolbox for the next upgrade.

Daniel Overby Hansen

Room D Thu 28 May 2026 · 11:05 - 11:50

Trash Talk - Optimising Java Garbage Collection

Understanding how Java organizes memory is important for every Java professional, but this particular topic is a common knowledge gap for many software professionals. This talk will help you understand the different parts of Java memory and the various garbage collectors so you can select your preferred one.
Lets explore how memory management can help you effectively improve performance.

Željko Kozina

Room E Thu 28 May 2026 · 11:05 - 11:50

What to consider when building resilient and efficient system

In this talk, we’ll dive into what it really takes to build resilient, efficient systems both on-prem and in the cloud. Too often, teams prioritize “just make it work” over thoughtful design, leading to brittle architectures and rising costs.

We’ll explore concepts like load shedding what it is, when to apply it and whether failing fast is always the right strategy. Is horizontal scaling the answer to every problem? Not necessarily. Sometimes it simply masks deeper issues.

Using a simple, relatable architecture as an example, we’ll break down how different types of requests behave, how to avoid wasting CPU cycles, and why certain architectural decisions are key to building systems that are both reliable and cost-effective.

Matej Nedić

Room F Thu 28 May 2026 · 11:05 - 11:50

Create and connect to your own MCP server in Java, with Quarkus

Model Context Protocol (MCP) has lately gained a huge amount of traction. But, as a Java developer, do you know how to easily create your own MCP server and then connect your AI-infused application to it?
Let's face it, many MCP servers nowadays are written in Python or TypeScript. Let's show that Java doesn't have to stay behind!
In this demo-focused session, I will demonstrate two concepts: how to build an MCP server and how to plug it into your application. For both parts, I will use Quarkus, because both these use cases have first-class support from Quarkus.

Jan Martiska

Room B Thu 28 May 2026 · 14:00 - 14:45

THE (ape)X FILES – The truth is out there – in the builder

In your daily work, like me, you might come across APEX applications that show really strange behavior. Instead of blaming this on supernatural phenomena or the work of aliens, you try to find a logical explanation for the behavior. Like a real investigating agent you explore the APEX Builder, looking for clues, on the hunt for the settings or components responsible for this behavior.
In this interactive session I will present the audience with a number of striking real world examples, and take them on a quest through the APEX Builder for the cause of the strange behavior. We will explore Builder settings you might not even know existed, Dynamic Actions that can be a source of confusion, and APEX API’s that sometimes give unexpected results. In the end we should be able to solve the mystery for each (ape)X file.

André van der Put

Room A Thu 28 May 2026 · 14:00 - 14:45

Analytics to Agents: A Survival Guide

The dashboards you've lovingly crafted aren't going anywhere, but they're about to get some very capable colleagues. This session charts a practical path from traditional Oracle Analytics Cloud workflows to an agent-augmented future, where AI doesn't replace your analytics expertise but amplifies it. We'll explore how agents can query your semantic models, interpret results, and take action — all while you maintain governance and control. Whether you're cautiously curious or ready to dive in, you'll leave with concrete strategies for introducing agents to your Analytics workflows without breaking what already works.

Mike Durran

Room C Thu 28 May 2026 · 14:00 - 14:45

Troubleshooting Oracle Memory

Oracle memory is complex, and errors such as ORA-04030 and ORA-4031 are common. How should we configure the memory, and what should we do when things go wrong? How do we configure and balance the PGA and SGA to best effect.

This talk should provide some insights into Oracle Memory usage and finding the source of memory issues.

Neil Chandler

Room D Thu 28 May 2026 · 14:00 - 14:45

Beyond the Prompt: Hiring Senior Engineers in the 2026 AI Surge

By 2026, the "Take-Home Assignment" has reached a breaking point. With multi-modal LLMs solving standard Spring Boot challenges in seconds, technical screenings now often measure prompt fluency rather than engineering excellence. When every candidate produces "perfect" code, how do you distinguish the true architects from the sophisticated synthesisers?

This session introduces a "back-to-basics" hiring framework that bypasses high-level syntax to test first-principles logic. We explore a shift from API design tasks toward hardware-inspired visual programming challenges, focusing on transistors, logic gates, and bitwise operations. By stripping away the abstractions AI has mastered, we create a "logical stress test" that exposes a candidate’s fundamental reasoning, spatial awareness, and depth of problem-solving.

In this session, we will cover:

The Evolution of the "Cheat": A 2026 update on how candidates use agentic AI to bypass traditional hurdles.

Designing for Reasoning: How to build "niche" logical frameworks and visual constraints that force first-principles thinking over pattern-matching.

The Judgment-First Framework: Shifting evaluation from "Does it work?" to "How do they verify?"—testing the ability to audit AI and catch architectural hallucinations.

Defining the 2026 Senior: Why seniority is shifting from Implementation Velocity to Architectural Integrity.

Martin Ladecký

Room E Thu 28 May 2026 · 14:00 - 14:45

Building Agentic AI Workflows with xAI on Vertex AI: From Code to Action

In this session, I will dive into agentic AI systems ; intelligent agents capable of reasoning, planning, and acting autonomously and demonstrate how to implement them end-to-end on Google Cloud’s Vertex AI. Attendees will get hands-on examples of:

Designing decision-making pipelines for autonomous AI agents using Vertex AI Workbench and Vertex Pipelines.

Integrating explainable AI (xAI) techniques to ensure transparency, interpretability, and trust in agentic workflows.

Writing and executing Python code for AI agents that interact with APIs, analyze data, and make automated decisions in real time.

Leveraging Vertex AI models for predictive tasks, RAG-based knowledge retrieval, and continuous learning loops.

This talk is highly technical and includes live coding examples to show exactly how agentic AI can be deployed responsibly at scale. By the end, participants will not only understand the architecture of agentic AI systems but also be ready to implement their own explainable, cloud-powered autonomous workflows.

Mohamed Mortadha Manai

Room F Thu 28 May 2026 · 14:00 - 14:25

How we did Analytics, and you should too.

In short, for the past year, my team at Celtra built a Data Pipeline, consisting of Open Source and Closed services, and proprietary solutions too.

As development evolves, we must keep pace. We will share our best practices in AI-assisted development and agentic workflows that helped us reach our goals.

Celtra is a digital advertising company, that establishes memorable connections with audiences and drives measurable results via impactful creatives. In order to drive those results and power our AI-efforts, we need to tackle the distributed data problem.

The challenge was ingesting data from external distribution systems or Advertising ecosystems (e.g., Facebook, Google) and then aggregate, transform, and serve it in report-ready formats for downstream stakeholders.

In raw numbers, this involves ingesting billions of rows of data, creating hundreds of data models, and serving it with minimum latency. To do so, we had to first understand the distributed data and the behaviour of external systems (e.g., rate limits) and then transform it too. All had to be performant, robust, and scalable.

We did the research on Data Governance and Lineage, and compared the tools:
- Infrastructure - Terraform (IaC)
- Ingestion - FiveTran, AirByte
- Storage - Snowflake
- Transformation - DBT, SQLMesh
- Analytics Architecture

This shapes into a very interesting topic, touching on the ever more important topic of gaining control and understanding of data flowing in and out of your company.

Ariel David Jančar

Room F Thu 28 May 2026 · 14:30 - 14:55

Losing the Joy of Tinkering: How AI Changes the Way We Think

I used to spend hours wiring Raspberry Pis, configuring Linux systems, and breaking networks just to rebuild them again. It was messy, hands-on, and deeply satisfying. Now, I type a few words into Cursor CLI and everything just… works. It’s magic—and somehow, a little tragic.

In this talk, we’ll explore how AI automation changes not just our workflows, but our brains. When we stop tinkering, do we stop thinking differently? Are we trading creativity for convenience, mastery for magic? Join me for a reflection (and a gentle provocation) on what we gain—and what we quietly lose—when machines start doing the fun parts for us.

Krisztián Papp

Room F Thu 28 May 2026 · 15:00 - 15:25

OpenSearch: The Open Source Path to Search and Observability

OpenSearch has become a cornerstone of open source search and observability, empowering developers and organizations to derive meaningful insights from unstructured data at scale. With OpenSearch officially joining The Linux Foundation in 2024, backed by tech giants such as IBM, SAP, Uber and AWS, it further cemented its position in the open source ecosystem.

In this session we’ll introduce OpenSearch, from indexing and analyzing unstructured logs to full observability capabilities across tracing, monitoring and security. We’ll share latest improvements in query performance and scalability, and real-time analytics, as well as its expanding ecosystem with new plugins and SDKs in multiple programming languages, and its compatibility with cloud-native environments. You’ll even find vector search and natural language processing capabilities for your AI/ML development.

Join us to hear it right from OpenSearch Ambassador and lead of the Observability TAG (technical advisory group) and discover how OpenSearch can fit into your observability architecture.

Dotan Horovits

Room B Thu 28 May 2026 · 15:10 - 15:55

APEX, MLE, and EXIF - Real JavaScript Magic detects fake images

What if your low-code application didn’t just think “File uploaded” when receiving an image, but instead: “Ah, sunset, 6:42 PM, somewhere near Hanover, taken with a Canon.” Or maybe even something like "AI generated image, possibly a. deep fake".
That’s exactly what becomes possible — with JavaScript running inside the database, powered by Multilingual Engine (MLE).
In this session, we’ll show how freely available JavaScript libraries can be used directly within the Oracle database to extract EXIF metadata from images — right at upload time, without external services. MLE provides the logic, your low-code platform handles the UI, and together you get an application that doesn’t just store images, but understands them.
The solution is lightweight, reusable, and offers plenty of room for your own ideas. Perfect for anyone who wants to go beyond simple drag-and-drop in low-code — expect real code, creative approaches, and that special moment when images turn into meaningful data.

Martin Bach

Sonja Meyer

Room A Thu 28 May 2026 · 15:10 - 15:55

Oracle GoldenGate 26ai: The Next Generation of Real-Time Data Streaming and AI Integration

For over 30 years, Oracle GoldenGate has led real-time data replication. With 26ai, it enters a new era: introducing major changes, deprecations, and innovations that now extend to AI, replicating vector data, powering LLMs, and enabling distributed AI. Learn what changed, what’s gone, and how to prepare and migrate.

Gilson Martins

Room C Thu 28 May 2026 · 15:10 - 15:55

What Building a Filesystem Taught Me about Oracle ASM

Oracle ASM abstracts away storage complexity — but how much of what's underneath can you actually get to, and what can you build when you try?

While building ASMFS (https://github.com/usrecnik/asmfs), an open-source project written in Rust that exposes ASM files via FUSE, I found out. Some of what ASM exposes is documented. Much of it isn't — like x$kffxp, the undocumented fixed table that maps extents to physical disk locations, or the file header transformation that only revealed itself by flipping bytes until something worked.

We'll explore how ASM organizes files — from the alias hierarchy that creates familiar directory structures and filenames, down to how extents are mapped to physical disk locations. We'll also compare two access methods: direct block device reads versus the DBMS_DISKGROUP package, and what each reveals about how ASM thinks about I/O.

Along the way, we'll cover how FUSE made it possible to expose all this as a mountable Linux filesystem — and what you discover when you push Oracle further than the documentation goes.

Urh Srečnik

Room D Thu 28 May 2026 · 15:10 - 15:55

We Like Things Predictable. So Does AI.

Even the best chefs burn a steak sometimes. Children never do exactly what you expect. And AI? Same prompt, different answer - every single time.

You can't make any of them predictable. And maybe that's okay.

Here's what I've been thinking: you don't succeed with children by making them predictable - you give them a stable home. You don't succeed in the kitchen by eliminating surprises - you keep it clean, your ingredients fresh, your knives sharp.

What if AI is the same? What if the answer isn't making AI deterministic, but giving it solid ground to stand on?

I've been exploring this question at Flox. We spent years building reproducible environments - locked dependencies, immutable builds, deterministic tooling. We did it for ourselves, for our sanity, for "works on my machine." But lately I'm wondering: did we accidentally build exactly what AI needed?

This talk isn't about answers. It's about a question that keeps nagging me: when everything around AI is predictable, does AI become... even more usable? I think the boring work of reproducibility might be the most exciting thing for AI.

Rok Garbas

Room E Thu 28 May 2026 · 15:10 - 15:55

From AI to Agent with Langchain4J

You’ve built AI features into your application. Models are wrapped in services, RAG is in place, tools are wired, and calls are flowing. Then requirements evolve. A single response is no longer enough. You need steps that follow each other, branches based on decisions, retries when things fail, and sometimes multiple actions happening at the same time. At that point, it becomes unclear how to structure the logic without losing control or readability.

In this talk, we move from AI calls to agentic systems using LangChain4j. We’ll explore common workflow patterns such as sequential, loop, parallel, and conditional flows, along with error handling as a first-class concern. From there, we’ll zoom out to agentic concepts like shared state with an AgenticScope, goal-oriented agents, and pure agentic setups. We’ll also look beyond AI-only approaches, mixing in non-AI agents and human-in-the-loop interactions. All of this will be built live, step by step, to show how these ideas work in practice.

You’ll leave with a concrete understanding of when to use workflows and when to rely on agents, how to combine both without confusion, and how to keep state, goals, and responsibilities explicit. The goal is not to build smarter prompts, but systems that are easier to reason about, easier to evolve, and better aligned with real-world application needs.

Loïc Magnette

Room F Thu 28 May 2026 · 15:30 - 15:55

Ship Teams, Not Just Code: Debugging Team Motivation and Process Debt

The presentation focuses on building teams.

Key topics will include delegation (team leads, defining roles & responsibilities), motivating engineers (ownership, mission vs. money, transparency), and running focused meetings.

We'll go through 5 real-life examples of before mentioned topics and look at how the presenter (coming from an engineering background) failed & learned on the job.

Aleš Čadež

Room B Thu 28 May 2026 · 16:10 - 16:55

Observability for APEX Heroes: Why OpenTelemetry Will Change Your Life (and Your Apps)

Let’s be honest: when users say “your app is slow,” what they mean is “I want you to magically fix something without any actionable data.” And like the responsible developer you are, you do what every APEX professional does:
- open the Activity Monitor
- stare at logs
- swear a bit
- pray to the PL/SQL gods
- blame ORDS
- then try again

OpenTelemetry finally ends this madness.

With built-in support in Oracle APEX, you can now trace requests end-to-end — from browser click to SQL execution and all the weird stuff in between. Think of it as having X-ray vision for your application, except it doesn’t require gamma radiation or a tragic backstory.

In this session, we’ll explore what OpenTelemetry is (spoiler: open, standard, not another Oracle-only thing), why it’s the coolest addition to APEX in years, and how it turns debugging from a séance into a science. We’ll follow real traces, catch slow database calls in the act, confront laggy REST services, and build dashboards that make you look like a productivity wizard.

If you’ve ever wished you could answer “why is this slow?” without sweating, guessing, or summoning dbms_output, OpenTelemetry is your new best friend.

Richard Martens

Room A Thu 28 May 2026 · 16:10 - 16:55

No Slides, Just Streams: Real-Time Data Pipelines on the OCI AI Data Platform

Real-time data processing is essential for modern analytics and operational decision-making. In this live-demo session, we show how end-to-end real-time data pipelines can be built using the OCI AI Data Platform, unifying streaming, data processing, analytics, and AI.

The session is almost entirely demo-driven. Apart from a short introduction, there are no slides — only live, working pipelines and real data in motion.

Using live public data feeds from Transport for London (TfL), we demonstrate two scenarios. The first focuses on real-time arrival analysis, where streaming data is processed and analyzed using SQL and visualized in Oracle Analytics Cloud to deliver up-to-the-minute insights. The second scenario shows real-time visualization of bus movements, enriching live GPS data to monitor routes, positions, and delays on dynamic dashboards.

Although the examples use public transport data, the same architecture applies directly to industrial and manufacturing use cases, such as real-time data streaming from production environments. To round off the session, we also demonstrate how ML/AI capabilities can be added to streaming pipelines, for example through anomaly detection or predictive insights.

This session is aimed at Oracle professionals looking for a practical, hands-on view of real-time analytics on OCI — without slides, just streams.

Žiga Vaupot

Room C Thu 28 May 2026 · 16:10 - 16:55

namesti.sql

Nameščanje baznega dela informacijskega sistema običajno pomeni ustvarjanje shem, tabel, indeksov, PL/SQL paketov, sekvenc in drugih objektov, ki jih aplikacija potrebuje za delovanje. Vsak od teh elementov ima svoj življenjski cikel, vendar mora biti končni rezultat v vseh okoljih enak – v razvojni, testni, šolski in nazadnje v produkcijski bazi. Če želimo ponovljivost, si ročnega izvajanja SQL ukazov preprosto ne moremo privoščiti. Okolja morajo biti postavljena na povsem enak način.

Pogosto se zgodi, da do ciljnega okolja nimaš dostopa in namestitev izvaja DBA pri naročniku. Zato je ključno vprašanje, kako pripraviti namestitveni postopek tako, da bo ponovljiv, predvidljiv in odporen na razlike med okolji.

V vlogi DBA‑ja, ki skrbi za nameščanje baznega dela, smo pred približno petnajstimi leti oblikovali koncept, ki temelji na dveh krovnih SQL skriptah: namesti.sql in parameters.sql. Ti skripti predstavljata izhodišče vsake namestitve. Z nekaj preprostimi pravili in osnovnimi kontrolami se je ta pristop izkazal za dovolj robustnega, da ga nismo bistveno spreminjali vse do danes. V predstavitvi si bomo pogledali, kako je zasnova videti v praksi in kako se primerja z novejšimi orodji, kot sta Liquibase in Flyway.

Mitja Golouh

Room D Thu 28 May 2026 · 16:10 - 16:55

Scaling to Zero is Easy!

The hard part is waking up again!

When working with Kubernetes, it is trivial to find out when your workload is not getting any HTTP requests and is therefore redundant. It is easy to scale it down. The machinery to intercept requests to applications that are not present is where things get complicated.

When we were faced with this issue we didn't want to have proxies or service meshes to do that. Rather we went back to fundamental way of how Kubernetes works and planted a temporary handler, which upon reception of request would wake up the application.

In this talk I'll demonstrate how all that maze of disconnected but cooperating controllers works, how to (mis)use the Endpoint API and how to achieve scaling to zero and back up with no networking overhead and just plain Java and fabric8 kubernetes client.

Patrik Duditš

Room E Thu 28 May 2026 · 16:10 - 16:55

Before the Clock Runs Out: A Safe Escape from Java EE 8

Many enterprises still rely on Java EE applications running on Java 8, but its age is increasingly becoming a liability. Missing security updates, outdated libraries, and limited platform capabilities slow development and increase operational risk. Today, the question is no longer why to migrate, but how to do it for mission-critical systems before it is too late.

In this session we will share our experience modernizing business-critical banking systems, where we migrated existing Java EE applications to a modern Jakarta EE10 and OpenJDK21.

We will highlight the toughest challenges, including javax -> jakarta namespace shift, legacy code remediation, integration test updates, and application server specifics such as Payara. Describing how we minimized manual coding and achieved the transition to the new ecosystem by combining AI tools with automated refactoring.

Finally, we will discuss strategies for a safe rollout to the production environment.

This session is designed for anyone facing the burden of legacy codebases who needs a proven blueprint for the safe modernization of critical systems.

Leon Stanko

Room F Thu 28 May 2026 · 16:10 - 16:35

Bottlenecks All the Way Down: AI Solved Code Gen, Testing Is The Next Step for Kubernetes Developers

AI has made writing code cheap, but validating that code has is an issue that code gen doesn't help with. This talk examines two connected bottlenecks that now slow teams down, the growing fragility of the OSS security ecosystem and the long feedback cycles in Kubernetes development.

On the security side, AI generated slop is overwhelming maintainers and reducing trust in vulnerability signals. False positives, low effort submissions, and an overstretched CVE pipeline are creating a system where real issues are harder to identify and burnout is reaching a tipping point.

On the engineering side, AI speeds up code generation, yet developers still wait for CI/CD and staging environments to learn whether that code actually works with real dependencies. Validation, not writing, has become the new bottleneck.

This talk proposes a shift in focus, stricter validation requirements for OSS security and faster, more realistic validation loops for Kubernetes developers. Attendees will learn practical workflows and patterns that improve trust, speed up iteration, and help teams collaborate with AI in a way that produces safer and higher quality software.

Jake Page

Room F Thu 28 May 2026 · 16:40 - 17:05

Detecting Memory Leaks with JFR and JMC

Your heap is growing, GC keeps thrashing, and performance nosedives. Where did the memory go? Java Flight Recorder (JFR) and JDK Mission Control (JMC) can help you figure that out.

In this session, we will look into Java Flight Recorder and how it can help you figure out where the problem lies. We will run a Java program that leaks memory and use a JFR recording and JDK Mission Control to see allocation hotspots, GC pressure, and hopefully fix the issue.

Your takeaway is how to use JFR and JMC to spot hard-to-find problems in your application.

Marco Sussitz

Room B Thu 28 May 2026 · 17:10 - 17:55

MCP Supercharges APEX: Less Clicking, More Kicking — Straight to Git Push Done

Low-code is growing up. Modern teams expect Git, clean CI/CD pipelines, reproducible deployments and structured development processes — and this is exactly where APEX meets a new generation of tools. With MCP and AI-powered workflows, low-code becomes not only faster but also smarter and more resilient. This session focuses on live demos: APEX apps coded with APEXlang, versioned in Git, tested automatically, deployed through pipelines, and supported by intelligent language-model workflows. Instead of slides, you'll see how SQLcl, APEXLang, Podman and MCP turn low-code into a real development flow that cuts repetitive work, boosts quality and frees teams to build better solutions. We will also take a critical look at security aspects, discuss architectural implications, and show alternative approaches beyond MCP. Finally, the session highlights what is already possible today when integrating APEX with MCP-style protocols and intelligent workflows — and where the journey can realistically go next.

Martin Bach

Sonja Meyer

Room A Thu 28 May 2026 · 17:10 - 17:55

Building Enterprise AI Applications over Governed Data using Oracle AI Data Platform

Enterprise AI promises transformation, but ungoverned data delivers chaos. This session explores Oracle AI Data Platform (AIDP) as the foundation for building trustworthy AI applications that operate on curated, compliant enterprise data. We'll examine how AIDP's medallion architecture—bronze, silver, gold—creates the governed data pipelines AI agents need to deliver reliable insights rather than confident hallucinations. From the unified catalog that tracks lineage across your data estate to the workbench where data engineers and AI developers collaborate, you'll see how Oracle has assembled the building blocks for enterprise-grade AI. Attend this session for a clear understanding of where AIDP fits in your AI strategy and how to get started.

Mike Durran

Room C Thu 28 May 2026 · 17:10 - 17:55

Time for Timeouts

Sometimes you have to wait, but you can’t wait forever—so we have timeouts. In a modern tech stack, there are many kinds of timeouts, each belonging to a specific layer, from the operating system to database software. Most of these won’t be a problem until the day that particular layer becomes unavailable. For example, what happens to your database clients when you shut down one of the servers in a high-availability environment? This presentation explores the various types of timeouts you might encounter and what you can do to avoid unresponsive systems. It uses examples that will be familiar to Oracle database developers and administrators.

Øyvind Isene

Room E Thu 28 May 2026 · 17:10 - 17:55

Scripting on the JVM with Java, Scala, and Kotlin

This talk will explore use of JVM languages as scripting languages, replacing the Bash and Python scripts common throughout the industry. We will walk through live-coded demonstrations of how the JVM's benefits of performance, compile-time safety, and vast library ecosystem are advantages over traditional script platforms, but also how language verbosity, build tool overhead, and lack of convenient libraries hampers the efforts. Lastly we will demonstrate how script-focused tooling is able to smooth over some of those issues, simplifying build configuration and providing suitable libraries to make the JVM truly a world-class scripting environment as robust as any scripting language out there.

Haoyi Li

Room F Thu 28 May 2026 · 17:10 - 17:35

Out of the Box, Into the Stream: Reliable Events in Microservices

Building event-driven microservices sounds simple if we neglect data consistency of course.
So how could we reliably publish events when your business logic and database updates must remain atomic? This is where the Transactional Outbox Pattern comes into play.
We’ll begin with a familiar problem: ensuring reliable event publishing from microservices. We’ll take a quick look at why distributed transactions rarely live up to their promise, how local transactions actually behave, and where inconsistencies tend to appear.
From there, we’ll see how the Outbox Pattern has stood the test of time. We’ll walk through a classic Kafka-based outbox setup, then evolve it using Kafka Connect for CDC and event delivery. To finish, we’ll show how CDC with Apache Flink can be used to further optimize the pipeline, enabling lower latency and more flexible stream processing.

Dominik Žnidaršič

Room F Thu 28 May 2026 · 17:40 - 18:05

Your AI Code Reviews Are Missing the Point (And How to Fix It)

Most AI code review implementations focus on the wrong metrics—counting comments generated or code accepted rather than measuring developer velocity and code quality improvements. The real value lies in intelligent context integration and organizational learning at scale.
This talk examines successful AI code review deployments across engineering organizations, revealing four critical success factors: seamless integration with your existing development context (codebase, tickets, architectural decisions), specialized review guidelines that scale from 5 to 50,000 repositories, comprehensive observability to understand where AI adds value versus where it creates noise, and strategic human-AI collaboration patterns.
We'll explore real case studies of teams who've moved beyond basic linting to utilizing AI code review that understands business logic, catches architectural anti-patterns, and actually accelerate PR cycles. You'll learn when AI reviews shine, when they don't, and how to measure the difference.
This isn't about replacing human reviewers—it's about building an intelligent system that amplifies human expertise and reduces cognitive load where it matters most.

Yishai Beeri

Room B Fri 29 May 2026 · 09:00 - 09:45

Simplify and accelerate development of AI-powered applications

Organizations are trying to balance potentially significant development costs against the business value they can derive from generative AI initiatives.

Oracle AI Optimizer and Toolkit can help you rapidly develop proofs of concept. It speeds experimentation and evaluation, by integrating your data with any large language model (LLM) and any embedding model, providing full control over the model parameters, chunking strategy, and semantic search.

Oracle AI Optimizer and Toolkit allows you to easily explore retrieval-augmented generation (RAG) and agents in a no-code environment, with the full power of Oracle AI Vector Search to optimize the accuracy and safety of your AI project. Use the built-in evaluation framework to find the right models and parameters for your specific use case.

Time to go into production? Turn your experiment into a highly scalable service with the API server included or easily generate microservices code in Java (Spring AI) or Python (LangChain) to deploy as independent microservices. Integration with Oracle Database 23ai provides the ability to manage millions of chunks and embedding vectors, giving you the scalability to meet your demands.

Vili Tajnić

Room A Fri 29 May 2026 · 09:00 - 09:45

IN? EXISTS? JOIN? Just Ask the Oracle Optimizer!

When it comes to writing performant SQL queries, few topics spark as much debate as the use of subqueries with (NOT) IN or (NOT) EXISTS. Opinions in the developer community are diverse: while some swear by a single approach, others follow best practices – and yet others avoid subqueries altogether, replacing them with joins.
But how valid are these approaches today? With every new Oracle release, the optimizer becomes more intelligent – and many long-standing rules lose their relevance. Time, then, to put some of these SQL myths to the test.
This talk uses concrete examples to highlight the differences between subqueries, semi- and anti-joins, and traditional joins. It explains which query transformations – such as subquery unnesting or partial join evaluation – the Oracle optimizer applies in order to deliver the best possible result, regardless of the SQL form used.

Andrej Pashchenko

Room C Fri 29 May 2026 · 09:00 - 09:45

Dissecting Performance Tuning

I still see SQL “performance tuning” that consists of random hints, superstition, bad indexing, voodoo, and parameter hacking, all applied with complete confidence and limited understanding. It’s not tuning. It’s hope over understanding.

Performance Tuning is not an art. It's not magic. It is the logical application of mathematical rules for deterministic outcomes.

Lets dissect Performance Tuning and see what happens in the real world as we try to avoid unintended consequences and look at what you really need to do to for effective Performance Tuning.

Neil Chandler

Room D Fri 29 May 2026 · 09:00 - 09:45

Leading Through the Shift: Architecting the Future-Ready Engineering Org with AI

Artificial Intelligence is more than a new toolset—it's a seismic shift, requiring engineering organisations to rethink their foundations. For technical leaders, this transition isn’t about just adopting code assistants or ML models. It calls for a strategic, intentional redesign—becoming architects of future-ready teams. This presentation moves beyond the AI hype to tackle the complex challenge of embedding AI at the core. We’ll examine how AI transforms not only what we build, but how we build, forcing a re-examination of structures and leadership. We’ll address how to harness AI while building resilient, high-performing teams. The session breaks down three key pillars:
- People (upskilling talent, fostering psychological safety, evolving roles),
- Process (embedding AI ethically and securely into workflows, refining quality gates, adapting agile methodologies), and
- Potential (unlocking innovation beyond mere productivity gains, fostering data-driven decision making, tackling previously intractable problems).
Attendees will leave with actionable frameworks and new perspectives for leading their teams through this transformative era, ensuring they don't just adapt to AI, but actively shape an engineering future where both human ingenuity and machine intelligence thrive.

Aleksandar Stoisavljevic

Room E Fri 29 May 2026 · 09:00 - 09:45

Unpoly: Single Page Experience with Server-Side rendering

Modern JavaScript frontends offer seamless navigation, elegant transitions, stateful components, offline caches, and many other features. However, managing these complex frontends can be challenging and time-consuming.

Unpoly is a progressive enhancement library for HTML, which leverages AJAX calls to bring the state back to the server side while still delivering a rich client-side interface. By leveraging a few attributes and some HTTP headers, Unpoly simplifies front-end development without sacrificing end user comfort.

The talk will walk through a sample application built with Jakarta MVC framework and almost no javascript demonstrating the power and simplicity that Unpoly may bring to your frontend.

Patrik Duditš

Room F Fri 29 May 2026 · 09:00 - 09:45

Jakarta Agentic AI - Status and Future

AI agents are one of the most prominent developments in enterprise and cloud native computing in decades. They promise to fundamentally accelerate automation, and productivity for many industries. Agents leverage ML, NLP, LLMs and other AI technologies. Examples include self-driving cars, security monitors, stock monitors, code/application generators, health monitors, customer service agents, manufacturing robots, etc.

The Agentic AI specification provides a set of APIs that make it easy, consistent, and reliable to build, deploy and run AI agents on Jakarta EE. The technology aims to do for developing AI agents what Servlet did for HTTP processing, Jakarta REST for RESTful web services, or Jakarta Batch for batch processing. It defines common usage patterns/life cycles for AI agents, a minimal LLM facade to define dynamic agent workflows that can change at runtime and integrates with other Jakarta EE APIs like CDI, Validation, JSON Binding, Persistence, Messaging, and more.

This session gives an overview of the Jakarta Agentic AI spec, its current status and planned future development, followed by a hands-on demo of real-life use cases like detecting potentially fraudulent bank transactions or automatically documenting source code after a push was made to the Git repository.

Werner Keil

Room B Fri 29 May 2026 · 10:00 - 10:45

Oracle Private Agent Factory: your no-code, privacy-friendly, secure agents platform

The wish to create AI agents is at its maximum. To satisfy this, a number of platforms allow you to easily build AI agents with no code required. By connecting pre-defined building blocks, you can create advanced workflows with AI capabilities.

You may not only want to build AI agents from scratch but also benefit from advanced prebuilt agents able to easily answer questions based on a set of sources, or a data analysis agent able to extract insights from a database.

Often, privacy is a concern with such kinds of platforms: “Who can see my data?” or “Why should I send my data out to an unknown cloud service to have an AI agent?” and the common “I want to run this locally, because I don’t trust anyone.”

Oracle answered these needs and concerns, and more, by introducing the Oracle AI Database Private Agent Factory. Run it on-premises, on your own hardware, run LLMs locally, and your data is safe and stays private inside your own corporate network. “How much?” you ask, and the answer is “free”! Or, run it in the cloud, but keep full ownership and control.

Join me to discover this product, which could look familiar if you are an n8n or comparable products user, and see what it does, how it does it, and give you ideas of interesting future usages.

Gianni Ceresa

Room A Fri 29 May 2026 · 10:00 - 10:45

If you have a problem, if no one else can help, and if you can find them, you can use the profilers.

Most of our database applications have complex business logic, spread out over a multitude of packages, procedures, functions, triggers.
A process usually starts off with a nice clean procedure. But it will execute others, which will do DML, that fires triggers, executing functions, ...

So now you are tasked with solving the problem that some process is slow.
If it's because of one of the queries, you'll analyze what's wrong using execution plans, AWR etc.
But what if it's not? Sure, you can add logging to the procedure and find out the problem is in the call of another procedure. Then you can add logging to that, and so on.

With the profiler and the hierarchical profiler Oracle has two powerful tools to trace exactly what happens during execution of plsql.
With these profilers you will immediately know exactly how much every part of the code that is run contributes to the total runtime.
So, then you can focus on that line that uses up 75% of the runtime, rather than wasting time on rewriting that obviously suboptimal loop only to find out it gave you a 1% improvement.

Both profilers have been around for quite a long time now, still in my experience they are rarely used.
It's more common for me to hear "Profiler? What's that" than "Yes, it's part of my standard tool set".
So, this talk is meant to make you more familiar with the profilers and to convince you that it couldn't be easier to instantly know which part of the code to optimize for maximum gain.

Erik van Roon

Room C Fri 29 May 2026 · 10:00 - 10:45

What's new for developers in Oracle AI Database 26ai

Oracle AI Database 26ai offers a wealth of new features, the number is in excess of 300 at the last count. And it grows even larger if you skipped Oracle Database 21c...

The focus on the latest long term support release can clearly be defined as "improved developer experience". But what exactly does that mean? In this presentation, delivered without any slides, you'll get a glimpse at some of the most noteworthy ones, delivered as a pure demo: just a SQL prompt and a lot of fun working through annotations, data usage domains, server-side JavaScript, sessionless transactions and many more. Come join me for an interesting live demo and let's have some fun together learning about some of the coolest, most interesting new developer features. And since no current presentation would be complete without secure AI-driven development there will be some of that, too.

Martin Bach

Room D Fri 29 May 2026 · 10:00 - 10:45

Your Code Base as a Crime Scene

The objective of Cluedo is to deduce how, where, and by whom the victim was murdered based on clues gathered from the other players. Colonel Mustard, spanner, Conservatory? Mrs. White, candlestick, Lounge? Ask questions, gather facts, and deduce answers before anyone else to win.

An objective look at your code base finds many problems: lost architectural vision, questionable code structure, security through obscurity, undeleted dead code, inconsistent coding paradigms, copy-pasted code, and more! Instead of considering current state, let's look at the journey. Did a sleep-deprived engineer misunderstand Stack Overflow, did the Product Manger making unreasonable commitments, did a non-technical leader inflict her technical decision? Causes are distributed throughout the org with plenty of blame to share.

Join me for a fun discussion about what triggered a decline of your code quality and what actions you, as the engineer, can take to protect the code and yourself!

Scott Sosna

Room E Fri 29 May 2026 · 10:00 - 10:45

From Git Panic to Git Zen: With a Little Help from AI

Everyone knows how to push, pull, and commit in Git. But Git’s advanced features like reflog, bisect, and interactive rebase are powerful — but often intimidating or unknown to many developers.

In this talk, you will:

Learn the core concepts and practical uses of reflog, bisect, and interactive rebase through clear, real-world examples.
Understand how mastering these features can solve common Git problems and improve your workflow manually.
Discover how AI integration with a custom MCP server backend automates these complex Git operations, reducing errors and saving time.
Experience a demonstration of human-AI collaboration that transforms Git Panic into Git Zen.

If you want to move beyond basic Git usage and explore how AI can help you work smarter, this session is for you.

Pasha Finkelshteyn

Room F Fri 29 May 2026 · 10:00 - 10:45

AI in Java: We've built a Server-side enterprise MCP server for Jakarta EE developers.

Java has seen a meteoric rise in AI through various AI agent libraries, such as LangChain4J and Embabel.
We've seen various platforms adopting MCP, such as Quarkus and Helidon, but these require running the MCP server on-premise.
But what if developers prefer the advantage of deploying an MCP service and exposing it as an HTTP service, just as they currently deploy their enterprise applications?

In this talk, I showcase an enterprise implementation of the MCP server that is fully powered by Jakarta EE features.
Developers will be introduced to MCP features (such as tools, prompts, resources and resource templates, logging, etc.), and how simple it is to integrate the MCP into their existing enterprise service application, with minimal configuration.
These features can be deployed to any Jakarta EE server of choice, using their existing deployment method of choice.

This talk will be beneficial for existing enterprise Java developers, as it will showcase the barrier to entry to learn such features to be minimal, and provide a little learning curve and have little impact on integration to existing business services, ensuring that developers can build AI applications with ease.

Buhake Sindi

Room B Fri 29 May 2026 · 11:10 - 11:55

Brihtno backupiranje virtualk

Inkrementalni backupi brez bitmap indeksov

Virtualne sisteme je najbolje varovati s shranjevanjem celotnih diskovnih slik. Količina podatkov pa nas hitro prisili v optimizacijo. Inkrementalni backupi so industrijski standard, njihova šibka točka so pogosto bitmap indeksi (CBT). Ti se ob rutinskih posegih – ugašanje strojev, razširjanje diskov in nekaterih drugih administrativnih posegih – radi invalidirajo ali izgubijo, kar vodi v dolgotrajne in potratne ponovne polne backupe.

Kaj se boste naučili na predstavitvi?
• Inkrementalni backup brez bitmapov: Spoznajte metodo, ki je odporna na sistemske posege in zahteva le eno začetno polno kopijo.
• Enostavna verifikacija: Način nudi enostavno verifikacijo, ali sta produkcijska in shranjena kopija identični, kar pogosto pri inkrementalnih backupih ni enostavno.
• Paradoks zmogljivosti: Ponavadi izvajanje backupa upočasnjuje produkcijski sistem. Spoznajte metodo, ki sistema ne upočasni, temveč ga pogosto celo občutno pohitri.

Predstavitev se osredotoča na ekosistem KVM (Oracle Virtualization - OLVM, oVirt, Proxmox, libvirt, ...), postopki pa so splošno uporabni tudi v drugih virtualizacijskih okoljih.

Sergej Rožman

Room A Fri 29 May 2026 · 11:10 - 11:55

Going Top-Down: A Deep Dive into Oracle Top-N Queries

Top-N queries are among the most established patterns in Oracle SQL, yet their implementation has evolved significantly over time. Early Oracle versions relied primarily on ROWNUM, later analytic functions provided more flexibility, and since 12c the SQL-standard row-limiting clause (FETCH FIRST …) has offered a clearer and more expressive alternative. With Oracle 26ai, the new QUALIFY clause adds yet another elegant option for specific use cases.

But how do all these approaches actually work under the hood?

In this deep dive, we take a detailed look at the different Top-N techniques available in Oracle. Using practical examples and live demos, we examine whether each method can reliably cover all requirements. We analyze the transformations applied by the optimizer, study the resulting execution plans and learn how to interpret their steps. We also explore the impact of indexes on these plans. This leads to the key question: are all Top-N variants equally performant, or do the differences go beyond syntax and readability?

The session provides a solid foundation for making informed decisions when designing Top-N queries in modern Oracle databases.

Andrej Pashchenko

Room C Fri 29 May 2026 · 11:10 - 11:55

007 Myths about SQL Performance

Many developers don’t pay attention to database performance when implementing new applications. The functionality of the application is important, but the efficiency of the SQL queries is often ignored. The DBA (or the query optimizer) does this job later. If necessary, we can still create some indexes, add hints or adjust some database parameters. And in the Cloud, the Oracle database does everything automatically and autonomously, anyway.
Really? Unfortunately the reality looks different. Many performance problems are caused by a poor software design, unsuitable data models or inefficient SQL statements. Correcting such issues at the end of a project or when the system is already in operation is often very costly and cannot be fixed with just a few configuration parameters.
With SQL live demos (all based on James Bond characters and movies), multiple choice questions and some funny stories that happened to me in my job as an Oracle consultant, I will explain why the impact of performance is important in every phase of a software project. You can check yourself what you know about SQL performance tuning and which myths you still believe in.

Dani Schnider

Room D Fri 29 May 2026 · 11:10 - 11:55

JavaScript? No. Java Scripts! — Discover Java as a Scripting Language

For years, scripting has been the domain of Python, Bash, or JavaScript — but if you’re like me, you were never quite comfortable with any of them. You can’t remember the Bash syntax, struggle with Python’s dependency drama, or accidentally turn 5 + 3 into "53" in JavaScript.

Luckily, modern Java is quietly stepping into this space. With the latest features, Java is no longer just a verbose enterprise language — it can now be lightweight, interactive, and even script-friendly.

In this talk, we’ll explore how Java has evolved into a practical scripting language. We’ll discover how the latest features — like shebang support, compact source files, and instance main methods — pair perfectly with records and text blocks to enable real-world automation.

For more advanced scripts, we’ll dive into handling dependencies and solving common scripting challenges using JBang. Through live demos, you’ll learn how modern Java features allow you to quickly prototype, automate tasks, and write scripts that are both powerful and maintainable.

By the end of the session, you’ll see Java in a new light: not just for large applications, but as a modern, practical scripting language.

Jens Knipper

Room E Fri 29 May 2026 · 11:10 - 11:55

Scotty I need Warp Speed - Ways to improve JVM startup

When running Java in the cloud, the JVM startup time can become a limitation. Especially when you look at spinning up new instances of an app as response to changes in load, the warmup time can be a problem. A couple of years ago you only had CDS that could help you with the startup but meanwhile there are a couple of solutions that can help you to let you JVM start super fast. These days we have CRaC, Native Image, Leyden, InstantOn, Ready Now, JITServer and Cloud Native Compiler, which are all solutions that help to reduce the startup time of a JVM. Each of those is not a Silver Bullet, so you have to choose the right solution for your problem.
This session will give you a short overview of the different solutions to increase your JVM startup performance and will try to give you some hints on how to choose the right solution for your application.

Gerrit Grunwald

Room F Fri 29 May 2026 · 11:10 - 11:55

From web shop to beverage AI assistant - rethinking the UX with AI

What happens when you turn a well-established Spring Boot web app with a React frontend and PostgreSQL backend into a smart Beverage AI Assistant? That’s precisely what we’re doing live: we give the monolith an agent upgrade via an MCP facade, let an agent orchestration service handle recommendations, shopping cart, and shipping, and enrich the LLM with local brewery knowledge using Agentic RAG. The trickiest part: state handling between the classic form flow and the dynamic LLM context. How do you make sure the chat and the shopping cart don’t talk past each other? It’s right at this interface where you’ll find the most delightful bugs, the cleverest workarounds, and the biggest aha moments.

Ingo Düppe

Room B Fri 29 May 2026 · 12:05 - 12:50

Oracle Večoblačnost (Multicloud)

Multicloud pristop omogoča organizacijam, da združujejo najboljše oblačne storitve različnih ponudnikov in tako izkoristijo prednosti vsake platforme.
Konkretno ,Oracle Multicloud omogoča uporabo najbolj pogostih OCI storitev - to so OCI bazne storitve in storitve povezane z Oracle bazo podatkov - neposredno v Azure, Google in AWS oblaku, ob tem, da administracija še vedno ostane v OCI.
Na predavanju bodo predstavljeno koristi Oracle Multicloud, obračunavanje strošov, arhitekture in druge podrobnosti.

Robert Korosec

Room A Fri 29 May 2026 · 12:05 - 12:50

Why this f***** Oracle does not use my index? [*fabulous]

You created it. It’s beautiful. Perfectly chosen columns, well-thought-out structure, proudly listed in USER_INDEXES.
And yet, when it actually matters, the Cost-Based Optimizer (CBO) looks at your index, snorts with contempt, and chooses a brute-force Full Table Scan instead. Sounds familiar?
Join this session as we dive into the “brain” of Oracle to understand why carefully designed indexes are sometimes completely ignored. No sugar - coating and with a healthy dose of sarcasm, we’ll look at situations where the database knows better - or only thinks it does.
We’ll find out why size matters, why everyone lies especially statistics - whether proximity really matters, and how an identity crisis can affect performance.

Join us and help convince the optimizer that your index is, in fact, fabulous.

Monika Lewandowska

Room C Fri 29 May 2026 · 12:05 - 12:50

Integrating AI into Performance Tuning Tools

As authors of APPM and JAS-MIN, we'll share which AI capabilities we chose to implement in our Oracle monitoring tools—and why. The session serves as a practical case study: attendees will see real implementations and the reasoning behind our technical decisions, providing a reference point for building AI support into their own software.

We'll cover our Java-based MCP (Model Context Protocol) implementation for APPM, including core concepts such as tools, resources, prompts, and instructions. We'll then demonstrate JAS-MIN's AI-assisted AWR/Statspack interpretation, showing how we transform raw performance metrics into plain-language diagnostics.

Expect live demos and practical insights from two tool authors who've already navigated these choices.

Kamil Stawiarski

Urh Srečnik

Room D Fri 29 May 2026 · 12:05 - 12:50

Personalizing Your Random Numbers

Random number generators (RNGs) generate "supposedly" random numbers for use in your application or solution. However, most RNGs are algorithmic, "pseudo", that anyone who deeply understands the algorithm may be able to regenerate the "random" numbers. Often this isn't too concerning: go ahead, recreate the shuffled songs played by Sonos.

But when it's important, it's REALLY important: cryptographic keys, digital signatures, authentication challenges, statistical analysis, gaming and gambling. Different algorithms have different entropy: higher entropy leading to better randomness - assuming, of course, no back doors were injected! That said, even strong algorithms displaying high entropy _may_ have faults to exploit. Should you be concerned?

In this session we'll dig into RNGs, pseduo-RNGs and, for the truly concerned or paranoid, learn to create custom true RNGs that are impossible to hack. Hope you can join us!

Scott Sosna

Room E Fri 29 May 2026 · 12:05 - 12:50

Revealing the magic behind Java annotations

Annotations, introduced in Java 5 two decades ago, have revolutionised how we develop Java applications. Traditional frameworks like Spring and Hibernate rely on runtime annotation processing using reflection to perform tasks such as dependency injection or implement persistence operations.

Alternatively, the Java Annotation Processing API, introduced in Java 6 in 2006, allows developers to hook into the compilation phase to process sources and react to annotations present on them. This was leveraged by libraries like Lombok and Dagger, and frameworks such as Micronaut, although their way of processing annotations varies. Other frameworks like Quarkus perform the annotation processing during build time, instead of the compilation time.

In this session, Micronaut Framework committer Álvaro Sánchez-Mariscal will explain with examples the different techniques that can be used for both runtime and compile-time processing of annotations, revealing the magic behind popular open-source projects.

Álvaro Sánchez-Mariscal

Room F Fri 29 May 2026 · 12:05 - 12:50

A2A in 45 Minutes: Building Interoperable AI Agents

Until recently, AI agents were advancing in intelligence but remained isolated. Developers attempted to make them collaborate by repurposing protocols like MCP, which connects language models to external tools and data. However, MCP was never intended for agent-to-agent communication. This gap became clear as agents grew more complex and collaboration became essential. In April, Google and its partners announced the A2A protocol, an open standard specifically designed to let agents discover each other, communicate, and collaborate across ecosystems.

This talk introduces the A2A protocol and shows how it enables agents to work together in multi-agent, multi-modal, and even long-running scenarios. We will explain its core principles, such as open standards, secure communication, and capability discovery through Agent Cards. You will also see how agents negotiate tasks, exchange context and artifacts, and coordinate their user experiences. We will highlight how A2A complements protocols like MCP by enabling agents to orchestrate one another while still using MCP internally to access tools and data.

By the end of this session, you will understand what problems A2A solves, how it fits into the broader agent ecosystem, and how it improves on previous approaches. You will leave with a clear understanding of how to design agents that can truly interoperate and contribute to the emerging Internet of Agents.

Loïc Magnette

Room B Fri 29 May 2026 · 14:00 - 14:45

Deep Dive: The Gemini CLI Oracle Extension for Developers and DBAs

Looking for the fastest way to integrate powerful AI into your Oracle database administration and development workflow? This introductory session is your hands-on guide to getting started with the Gemini CLI—the open-source command-line tool that brings Google's advanced AI models directly into your terminal.

We focus on immediate, practical application with zero prior AI experience required. You will learn:

Quick Setup: Install, configure, and authenticate the Gemini CLI in minutes.

Essential Commands: Discover the core commands to accelerate daily tasks. We’ll show you how to use natural language to quickly generate complex SQL queries, draft necessary infrastructure code (like Terraform or YAML), and explain unfamiliar log messages or error codes.

Context in Action: See how the CLI intelligently uses your local project files and context to provide accurate answers—making it a true, in-terminal assistant for troubleshooting and deployment.

Join us to transform your command line into an interactive, productivity-boosting tool, simplifying complex database and infrastructure tasks.

Kiran Tailor

Room A Fri 29 May 2026 · 14:00 - 14:45

I know what You did last Summer… and Winter… and Autumn… and Spring

Data auditing is extremely useful if You want to keep track of what was going on with the data in Your database. In this session I would like to touch on this topic from various perspectives such as why to implement data auditing at all, different approaches to data auditing, benefits of data auditing, various implementation methods, performance issues and data storage. I would also like to present “Data JAM for Oracle” - an Oracle database auditing utility I’ve built - as one potential solution for data auditing.

Zoran Tica

Room C Fri 29 May 2026 · 14:00 - 14:45

The Great Unknown. All about NULL in the Oracle Database

"NULL: three letters that can perplex even seasoned Oracle professionals."
Well, if that's ChatGPTs idea of an engaging title, I guess I'll have to do this the hard way!
Come learn (a little) about the history of NULL, and how best to deal with it as either DBA or Developer.
Aimed as an introductory course, but I suspect most will learn something new (for sure I did when preparing it). Worst case you get to feel smug, and enjoy some pictures of animals, some of them alive!

Patrick Jolliffe

Room D Fri 29 May 2026 · 14:00 - 14:45

Challenging the Python-First Mindset: ETL and Data Migration with Java

When it comes to data analysis and ETL pipelines, Python and its data science ecosystem are often considered the default choice. In this talk, we challenge that assumption.

When our company was tasked with implementing a data migration platform, we deliberately chose Java as our primary language, leveraging its battle-tested enterprise tooling, strong database integration, and mature testing ecosystem.

We will present how we built an ETL platform to migrate a heterogeneous landscape of business-critical applications while keeping all systems operational and bidirectionally synchronized using several CDC techniques.

Java enabled a balanced combination of domain-driven object-oriented design and functional paradigms, which we used for expressive and testable data transformations. We will walk through the application architecture, including middleware components for data transformation, validation, and migration reporting. Integration testing with JUnit and Testcontainers played a key role when connecting multiple input and output systems. For scalability and orchestration, we combined this with modern cloud-native technologies such as GCP, Apache Spark, serverless Dataproc, and Airflow.

The talk concludes with an honest assessment of where Java excels, where it struggles, and whether it is a viable choice for data analysis and large-scale data migration projects in 2026.

Lukas Berger

Room E Fri 29 May 2026 · 14:00 - 14:45

Netflix dropped REST, but should you? REST developers' Assessment of Spring GraphQL

GraphQL is undoubtedly one of the popular kids on the Development playground right now. Last year, Netflix announced that it was dropping REST completely in favour of GraphQL. GraphQL is a new way to build APIs, one that's easier, faster, better, and more flexible than the old RESTful API system, or at least that's the promise. In this session, we will cover the basics of Spring GraphQL, explore its strengths and weaknesses, and determine whether it is a better option for your company than a conventional REST API.

Željko Kozina

Room F Fri 29 May 2026 · 14:00 - 14:45

Flox + Kubernetes: Be Deploying, Not Building Images

Building container images works. Millions of teams do it every day, and that's great.

But what if there was another way?

I've been experimenting with deploying workloads to Kubernetes without building images at all. Using Flox, you define your environment once, and Kubernetes runs it directly - no Dockerfile, no registry push, no waiting for builds. The environment is already reproducible, so you just... deploy it.

This talk shares what we've learned from this approach: faster iteration cycles, simpler CI pipelines, and SBOMs that come for free. It's not for everyone, and it's not replacing what you're doing. But as Ted Lasso would say: "Be curious, not judgmental." This might give you some ideas.

Rok Garbas

Room B Fri 29 May 2026 · 15:05 - 15:50

The Love and No Love with Oracle Performance Tools

Oracle Database provides a solid set of performance tools that are essential for troubleshooting and optimization. While they are powerful and useful, some aspects are less than ideal. In this talk, I’ll go over the features that work well for me, the ones that don’t, and the changes that, in my view, would make these tools significantly better.

Timur Akhmadeev

Room A Fri 29 May 2026 · 15:05 - 15:50

The Rise of the Invisible DBA: AI and Oracle

In this session, I’ll show how AI can work as an invisible DBA for Oracle databases. The idea is not to replace DBAs, but to automate the boring and repetitive parts of the job. Step by step, I’ll demonstrate how to use MCP with VS Code to connect AI to an Oracle database and run common DBA tasks like basic health checks, performance analysis, and troubleshooting. Everything will be done using simple, practical examples, focusing on how MCP helps keep the interaction structured and safe. By the end, you’ll have a clear idea of how AI can quietly assist in day-to-day Oracle DBA work without getting in the way.

Rodrigo Jorge

Room C Fri 29 May 2026 · 15:05 - 15:50

The Bank Crash Chronicles: Why Our 11TB Oracle Exadata Architecture Failed

When dealing with 11TB of data and 10 billion records, the "standard" ways of ensuring data integrity begin to fail. Conventional read-side duplicate checks can become a bottleneck that can paralyse a system, especially under the scrutiny of financial regulators such as BaFin.

This session moves beyond the theory of database constraints and provides a battle-tested blueprint for managing massive data volumes in highly compliant environments. We will explore the architectural shift from read-side validation to a Write-Side Integrity Strategy, detailing the logic, the specific Spring-based implementation, and the performance gains realised.
Furthermore, we will break down the ongoing "Exit Strategy" from legacy Oracle Exadata to Cloud PostgreSQL. We will share best practices for a Zero-Loss Migration, ensuring 100% data integrity is maintained as we move billions of records under extreme time and regulatory pressure.

Martin Ladecký

Room D Fri 29 May 2026 · 15:05 - 15:50

Infusing Java with AI: 5 Tips and Strategies for GenAI-Powered Applications

AI is transforming software development, and Java is no exception. But are you ready to move beyond simple API calls and truly integrate Generative AI into your applications? In this talk we will discuss 5 advanced strategies for building intelligent, resilient, and scalable Java applications powered by LLMs. From local LLM deployments to architecting complex AI workflows, mastering context management, and leveraging LLMs for powerful data transformation, we will help you discover how to leverage GenAI to solve real-world problems and gain a competitive edge. This isn't your average AI 101 – prepare to level up your Java skills for the age of AI. Prepare for the examples using Jakarta EE specs, LangChain4J and Quarkus.

Rustam Mehmandarov

Room E Fri 29 May 2026 · 15:05 - 15:30

Stop Refreshing the Dashboard: Real-Time Reporting with CDC and Apache Flink

Imagine needing the latest event updates delivered straight to your dashboard, tailored to your needs and available the moment something changes. With Apache Flink and some clever use of Change Data Capture (CDC) on PostgreSQL, our team set out to make that a reality.

In this talk, I’ll share our journey of combining Flink’s stream processing capabilities with real-time data captured directly from the application database. We’ll look at how CDC can be used to drive reports without relying on periodic batch jobs or constant dashboard refreshes.

We’ll cover practical topics such as filtering and shaping event streams for reporting use cases, enriching live data by joining it with historical datasets, and handling common challenges like late events and evolving schemas. Along the way, we’ll discuss architectural choices, including when a Lambda architecture still makes sense and when a simpler Kappa-style approach is enough.

The focus is on real-world trade-offs, lessons learned, and patterns you can apply when building fast and flexible reporting pipelines powered by streaming data.

Dominik Žnidaršič

Room F Fri 29 May 2026 · 15:05 - 15:30

Rebuilding the Plane Mid-Flight: Tackling Massive Tech Debt Without Slowing Down

How do you refactor 70% of a mission-critical system without halting feature delivery? In this talk, I’ll share the real-world story of how our team addressed technical debt at scale – transforming a legacy, tightly-coupled ad tech platform into a resilient event-driven microservices architecture, all while continuously delivering business value. We’ll delve into the technical strategy (domain-driven design, an incremental ‘migration approach, & an Architectural Runway that enabled feature development without big-bang rewrites). Equally important, I’ll cover the leadership moves that made it possible: how we built stakeholder buy-in for refactoring, fostered a culture of blameless post-mortems and innovation time, and mentored teams on new practices to elevate our code quality.
Attendees will learn practical techniques to tackle technical debt in large systems – from modularizing a monolith to improving observability for faster incident resolution. I’ll share the risks we managed, the mistakes we made (and learned from), and the quantitative outcomes we achieved (like improved uptime and a double-digit increase in platform capacity). Whether you’re an architect or team lead, you’ll gain insights on balancing fast delivery with long-term sustainability, and how to lead your team through complex technical change – without sacrificing velocity. This case study will leave you with concrete ideas & encouragement to turn your own technical debt into an opportunity for innovation.

Taimoor Alam

Room E Fri 29 May 2026 · 15:35 - 16:00

Debugging Class Loading with GDB

Have you ever wondered why a particular class was loaded at that exact moment? Getting to the bottom of it can be surprisingly hard.
I recently encountered a bug with a particularly nasty linkage error in which classes were loading unexpectedly, and I needed to figure out what triggered that. In this talk, I’ll walk through the steps I used to track it down. We’ll build a slowdebug JVM, attach GDB, and use breakpoints and inspection to identify the precise cause of the class loading.
If you’re curious about how the JVM really works and want to peek under the hood, this talk is for you.

Marco Sussitz

Room F Fri 29 May 2026 · 15:35 - 16:00

Pragmatic Java: A Return to Basics

Java is almost exclusively associated with large enterprise software, complex builds and heavyweight frameworks.

In this talk, we return to basics and look at Java as a practical language for writing small programs, scripts, and everyday utilities. We’ll explore what modern Java offers when you strip things down and show how to write useful Java programs without traditional build tools or project scaffolding. Along the way, you’ll see why this approach can be surprisingly productive for automation, data processing, prototyping, and one-off tasks.

Blaž Mrak

Room B Fri 29 May 2026 · 16:00 - 16:45

Real stupidity beats artificial intelligence every time.

People keep saying that large language models (LLMs) will completely change the IT world. Sure, they will! Programmers no longer need to memorize every weird syntax, library, or obscure function in Java, C#, or Python. With LLMs, you don’t have to master just one language—you’re free to choose the right tool for the job. But beware: artificial intelligence alone won’t save you if you don’t provide it with the right context. In practice, instead of spending your time buried in books that often get outdated faster than you can read them, it’s way smarter to focus on things that truly matter: understanding how programs work, how they communicate, what’s the difference between multithreading and multiprocessing, how shared memory actually behaves, and how to avoid painful mistakes involving I/O or networks. This may sound obvious, but in the era of LLMs, this approach is crucial. In this talk, I’ll show you why even the smartest AI will fail miserably if we don’t invest in building our own solid “internal context.” And I will show you what you can achieve if your internal context is strong and well structured - you will see how you can even build hacking tools like IMSI catcher - even if don't know right now what is it and why this is cool ;)

Kamil Stawiarski

Room A Fri 29 May 2026 · 16:00 - 16:45

Learning from Mistakes: My Biggest Failures in Database Projects

My professional career as a database developer and consultant is coming to an end – time for some self-critical reflection.
In 35 years of professional experience, I have seen many classic mistakes in database projects: from unclear requirements and questionable data models to terrible SQL queries. I have already spoken about some of these bad examples at previous conferences. Without mentioning the names of the customers, of course.
But what about my own mistakes? In this self-critical presentation, I look back on my most serious mistakes and embarrassing design errors that I caused myself or contributed to in various projects. What did I learn from these mistakes, and how would I do things better today?

Dani Schnider

Room C Fri 29 May 2026 · 16:00 - 16:45

Branch, Process, Validate: Decoding the APEX Execution Order, from URL to Commit

"Why is this item null? It should have a value!"

If you've ever said this you've hit the common wall of APEX execution flow confusion
Let's admit: the execution order of processes, validations, and branches is a common source of confusion, especially when one just started developing in APEX.
I invite you to join me for a practical, back-to-basics session dedicated to understanding the APEX engine's timeline.
We'll solve the "mysteries" of page execution:
- when computations and default values are set
- which process runs first
- why branches must execute last
- how to read the engine diary - APEX Debug Log - to instantly verify the execution order of every component on our page

After this session, you will no longer spend time debugging mysterious flow issues.You will simply know the APEX engine's next move.

Mirela Ardelean