top of page
Java Day Istanbul is one of the most effective international community driven software conference of Türkiye supported by Istanbul Java User Group. The conference helps developers to learn the newest technologies about Java, Web, Mobile, Big DATA, Cloud, DevOps, Agile and Future. Java Day Istanbul also helps developers, tech companies, and startups to establish a good network among them.
12
25
550+
60+
EDITION
SPEAKERS
COMPANIES
ATTENDEES
SPEAKERS
Get ready to join us in Istanbul to network with peers, learn about the latest industry trends, make new connections, and more #JDI26
Akın Kaldıroğlu
@Backend Guru
MasterClass: Hexagonal I Clean I Onion Architecture
MasterClass: Hexagonal I Clean I Onion Architecture
Lemi Orhan Ergin
@Craftgate
The Rebirth of Software Craftsmanship in the AI Era
The Rebirth of Software Craftsmanship in the AI Era
Dr. Naci Dai
@Eteration
Engineering Scalable AI with Java
Description: Python powers AI experiments—but production AI is a systems challenge. As agentic AI moves beyond notebooks into long-running, multi-agent platforms, the JVM re-emerges with strengths in concurrency, reliability, and safety. Discover why serious AI systems are increasingly built on the JVM.
Engineering AI at Scale: Why the Java Is the New Control Plane
Python might have won the battle, but Java will win the war for AI agents
For the past few years, the software industry has blurred two very different activities under the same label: training AI models and operating AI systems.
Because Python dominates the research and training phase, it has often been assumed incorrectly that it should also be the default choice for building production-grade, agentic AI platforms.
As organizations move from simple chatbots to stateful, long-running, multi-agent systems, a different set of engineering concerns begins to dominate: concurrency, orchestration, memory efficiency, type safety, observability, and supply-chain governance. These are not research problems they are systems problems.
We are at the beginning of the industrialization phase of AI, where experimental prototypes must be re-engineered into reliable platforms. In this phase, the JVM often overlooked in recent AI discussions re-emerges as a highly competitive runtime for agent orchestration and inference pipelines.
We will explore how recent and emerging JVM capabilities change the architectural landscape: how virtual threads dramatically simplify large-scale agent orchestration; how advances in JVM memory layout unlock efficient in-process vector workloads; and how type-safe agent frameworks such as LangChain4j help contain the failure modes of non-deterministic LLM outputs.
This is not an anti-Python talk. Python remains unmatched for experimentation and model development. Instead, this session focuses on what happens after the notebook when AI systems must run continuously, scale predictably, and fail safely.
Attendees will leave with a clearer mental model for choosing the right runtime for different phases of the AI lifecycle, and a practical understanding of why many production AI systems are increasingly being built on the JVM.
Engineering Scalable AI with Java
Description: Python powers AI experiments—but production AI is a systems challenge. As agentic AI moves beyond notebooks into long-running, multi-agent platforms, the JVM re-emerges with strengths in concurrency, reliability, and safety. Discover why serious AI systems are increasingly built on the JVM.
Engineering AI at Scale: Why the Java Is the New Control Plane
Python might have won the battle, but Java will win the war for AI agents
For the past few years, the software industry has blurred two very different activities under the same label: training AI models and operating AI systems.
Because Python dominates the research and training phase, it has often been assumed incorrectly that it should also be the default choice for building production-grade, agentic AI platforms.
As organizations move from simple chatbots to stateful, long-running, multi-agent systems, a different set of engineering concerns begins to dominate: concurrency, orchestration, memory efficiency, type safety, observability, and supply-chain governance. These are not research problems they are systems problems.
We are at the beginning of the industrialization phase of AI, where experimental prototypes must be re-engineered into reliable platforms. In this phase, the JVM often overlooked in recent AI discussions re-emerges as a highly competitive runtime for agent orchestration and inference pipelines.
We will explore how recent and emerging JVM capabilities change the architectural landscape: how virtual threads dramatically simplify large-scale agent orchestration; how advances in JVM memory layout unlock efficient in-process vector workloads; and how type-safe agent frameworks such as LangChain4j help contain the failure modes of non-deterministic LLM outputs.
This is not an anti-Python talk. Python remains unmatched for experimentation and model development. Instead, this session focuses on what happens after the notebook when AI systems must run continuously, scale predictably, and fail safely.
Attendees will leave with a clearer mental model for choosing the right runtime for different phases of the AI lifecycle, and a practical understanding of why many production AI systems are increasingly being built on the JVM.
Ahmed Enis Erkaya
@TÜBİTAK BİLGEM YTE
Yapay Zekâ Ekseninde Kod Gözden Geçirme: İyi ve Kötü Pratikler
Description: Bir kod gözden geçirmenin 10 dakikasını AI ile 3 dakikaya indirmek mümkün. Peki kaliteyi kaybetmeden? Bu oturumda gerçek deneyimlerden dersler çıkarıyoruz.
AI destekli kod gözden geçirme, yazılım geliştirme süreçlerini yalnızca hızlandırmakla kalmayıp geliştiricilerin problem çözme yaklaşımını da yeniden şekillendiriyor. Bu oturumda, yapay zekânın güçlü ve zayıf yönleri ele alınırken, günlük iş akışına nasıl etkili şekilde entegre edilebileceğine dair pratik yöntemler paylaşılacak. “Hata bulan” bir araçtan öte, geliştiriciye yol gösteren bir asistanın nasıl değer kattığı gerçek örnekler üzerinden incelenecek. Ayrıca kod gözden geçirmede iyi ve kötü pratikler, açık kaynak ve ticari araçlardan örneklerle değerlendirilecek; yapay zekânın sınırları, yanıltıcı veya eksik önerilerle başa çıkma stratejileri tartışılacak. Katılımcılar, hem teknik içgörüler hem de doğrudan uygulanabilir araç ve yöntemlerle, AI destekli kod incelemeyi ekiplerinde etkin bir şekilde hayata geçirebilecekleri net kazanımlarla oturumdan ayrılacak.
Yapay Zekâ Ekseninde Kod Gözden Geçirme: İyi ve Kötü Pratikler
Description: Bir kod gözden geçirmenin 10 dakikasını AI ile 3 dakikaya indirmek mümkün. Peki kaliteyi kaybetmeden? Bu oturumda gerçek deneyimlerden dersler çıkarıyoruz.
AI destekli kod gözden geçirme, yazılım geliştirme süreçlerini yalnızca hızlandırmakla kalmayıp geliştiricilerin problem çözme yaklaşımını da yeniden şekillendiriyor. Bu oturumda, yapay zekânın güçlü ve zayıf yönleri ele alınırken, günlük iş akışına nasıl etkili şekilde entegre edilebileceğine dair pratik yöntemler paylaşılacak. “Hata bulan” bir araçtan öte, geliştiriciye yol gösteren bir asistanın nasıl değer kattığı gerçek örnekler üzerinden incelenecek. Ayrıca kod gözden geçirmede iyi ve kötü pratikler, açık kaynak ve ticari araçlardan örneklerle değerlendirilecek; yapay zekânın sınırları, yanıltıcı veya eksik önerilerle başa çıkma stratejileri tartışılacak. Katılımcılar, hem teknik içgörüler hem de doğrudan uygulanabilir araç ve yöntemlerle, AI destekli kod incelemeyi ekiplerinde etkin bir şekilde hayata geçirebilecekleri net kazanımlarla oturumdan ayrılacak.
Alper Hankendi
@ Hepsiburada
System Design Is Not Just Engineering: The Anatomy of Architectural Pitfalls
Description: System design isn’t just coding. it’s about decisions made under pressure and shifting goals. At Hepsiburada, where systems serve millions, we learned failures otfen stem from choices, not code. I’ll share real pitfalls and steps to build resilient, lasting systems. I'll explain PEDALS method.
Every big system begins with a simple request. At Hepsiburada, serving millions of users, we learned that the biggest failures didn’t come from bad code, but from architectural decisions made under pressure and shifting priorities. In this talk, I’ll share real pitfalls we faced, the epic failures that followed, and the lessons they left behind.I’ll also introduce the PEDALS framework as a practical way to approach decision-making in system design. Because in technical architecture there is never a single right answer only trade-offs, choices, and their consequences. This session ties those hard-learned stories to a framework you can apply to design systems that last.
System Design Is Not Just Engineering: The Anatomy of Architectural Pitfalls
Description: System design isn’t just coding. it’s about decisions made under pressure and shifting goals. At Hepsiburada, where systems serve millions, we learned failures otfen stem from choices, not code. I’ll share real pitfalls and steps to build resilient, lasting systems. I'll explain PEDALS method.
Every big system begins with a simple request. At Hepsiburada, serving millions of users, we learned that the biggest failures didn’t come from bad code, but from architectural decisions made under pressure and shifting priorities. In this talk, I’ll share real pitfalls we faced, the epic failures that followed, and the lessons they left behind.I’ll also introduce the PEDALS framework as a practical way to approach decision-making in system design. Because in technical architecture there is never a single right answer only trade-offs, choices, and their consequences. This session ties those hard-learned stories to a framework you can apply to design systems that last.
Mihaela Gheorghe-Roman
@Systematic
Chaos meets code: Fractals as a playground for modern Java features
Description: Discover how modern Java relates to fractals and chaos theory! In this live-coded, interactive session, we’ll explore recursion, virtual threads, pattern matching & records to render fractals in real time - blending math, art, and code you can take home for your own projects.
Fractals are a unique blend of mathematics and visual art, and modern Java features provides powerful tools to explore them in new ways. In this talk, I’ll share how I’ve experimented with fractal generation using recent Java features - from virtual threads to enhanced pattern matching and records.
We’ll look at practical techniques for combining recursion with modern language and runtime capabilities to render fractals in real time, while reducing the traditional complexity of concurrency and boilerplate. You will leave with clear code examples and ideas you can adapt for visual, educational, or exploratory projects of your own.
Chaos meets code: Fractals as a playground for modern Java features
Description: Discover how modern Java relates to fractals and chaos theory! In this live-coded, interactive session, we’ll explore recursion, virtual threads, pattern matching & records to render fractals in real time - blending math, art, and code you can take home for your own projects.
Fractals are a unique blend of mathematics and visual art, and modern Java features provides powerful tools to explore them in new ways. In this talk, I’ll share how I’ve experimented with fractal generation using recent Java features - from virtual threads to enhanced pattern matching and records.
We’ll look at practical techniques for combining recursion with modern language and runtime capabilities to render fractals in real time, while reducing the traditional complexity of concurrency and boilerplate. You will leave with clear code examples and ideas you can adapt for visual, educational, or exploratory projects of your own.
Marco Belladelli
@IBM
Talk to Your Data: Hibernate-Powered Natural Language Data Access in Java
Description: Discover how Hibernate, the relational data library you know and love, meets the AI with the help of Quarkus world to make your database speak to you.
Explore how Hibernate ORM, Quarkus, and LangChain4j can join forces to enable natural language interactions with relational data in your Java application. Leveraging Hibernate’s deep understanding of your domain model and data access layer, new tools now bridge the gap between structured information and LLMs, allowing users and agents to query and explore data using plain language.
During this session we will demonstrate with a live demo of a Quarkus application how to interact in natural language with your existing Hibernate data model. In particular, we will show both how to manipulate data and how to use them to implement a simple but effective RAG pattern. This particular style of data interaction, including RAG, will have the advantage of leveraging Hibernate’s type-safe mappings, semantic-rich HQL, and fail-fast validation, making this data manipulation process more effective and less prone to hallucinations.
Finally, we will also showcase how to easily expose this tool as an MCP server through Quarkus, allowing it to be conveniently consumed in an agent to agent framework.
Talk to Your Data: Hibernate-Powered Natural Language Data Access in Java
Description: Discover how Hibernate, the relational data library you know and love, meets the AI with the help of Quarkus world to make your database speak to you.
Explore how Hibernate ORM, Quarkus, and LangChain4j can join forces to enable natural language interactions with relational data in your Java application. Leveraging Hibernate’s deep understanding of your domain model and data access layer, new tools now bridge the gap between structured information and LLMs, allowing users and agents to query and explore data using plain language.
During this session we will demonstrate with a live demo of a Quarkus application how to interact in natural language with your existing Hibernate data model. In particular, we will show both how to manipulate data and how to use them to implement a simple but effective RAG pattern. This particular style of data interaction, including RAG, will have the advantage of leveraging Hibernate’s type-safe mappings, semantic-rich HQL, and fail-fast validation, making this data manipulation process more effective and less prone to hallucinations.
Finally, we will also showcase how to easily expose this tool as an MCP server through Quarkus, allowing it to be conveniently consumed in an agent to agent framework.
Serkan Çavuşoğlu
@MongoDB
From Full-Text to Vector Search: Modern Search Architectures for Java Applications
Description: Search is no longer just about matching keywords. Today’s users expect applications to understand what they mean, not just what they type. In this talk, we will journey from classic full-text search to semantic retrieval, vector embeddings, RAG, and agentic AI patterns that are reshaping modern software.
Using a Java application perspective, we will explore how to build smarter search experiences with MongoDB Atlas Search, combine lexical and semantic approaches, and design architectures that power AI-driven assistants and intelligent applications.
This session is for developers and architects who want to move beyond traditional search and understand how modern retrieval systems are built in practice.
From Full-Text to Vector Search: Modern Search Architectures for Java Applications
Description: Search is no longer just about matching keywords. Today’s users expect applications to understand what they mean, not just what they type. In this talk, we will journey from classic full-text search to semantic retrieval, vector embeddings, RAG, and agentic AI patterns that are reshaping modern software.
Using a Java application perspective, we will explore how to build smarter search experiences with MongoDB Atlas Search, combine lexical and semantic approaches, and design architectures that power AI-driven assistants and intelligent applications.
This session is for developers and architects who want to move beyond traditional search and understand how modern retrieval systems are built in practice.
Anton Arhipov
@JetBrains
Spec-Driven Development with AI Agents: From High-Level Requirements to Working Software
Description: Spec-driven development brings structure—turning vague prompts into clear, trackable workflows. Learn how to guide AI agents with specs, plans, and tasks to produce reliable, reviewable code every time.
Spec-Driven Development with AI Agents: From High-Level Requirements to Working Software
Description: Spec-driven development brings structure—turning vague prompts into clear, trackable workflows. Learn how to guide AI agents with specs, plans, and tasks to produce reliable, reviewable code every time.
Haoyi Li
Modernizing the Java Build Tool Experience
The Java language is known to be performant, easy to use, and with great IDE support, but Java build tools like Maven or Gradle don't always live up to that reputation.
This talk will explore why build tooling is fundamentally such a difficult domain to work in, but also how common concepts from object-oriented programming have the potential to greatly simplify the build tool experience.
We will end with a demonstration of an experimental new Java build tool "Mill" that makes use of these ideas, proving out the idea that Java build tooling has the potential to be much faster, safer, and easier than it is today.
The Java language is known to be performant, easy to use, and with great IDE support, but Java build tools like Maven or Gradle don't always live up to that reputation.
This talk will explore why build tooling is fundamentally such a difficult domain to work in, but also how common concepts from object-oriented programming have the potential to greatly simplify the build tool experience.
We will end with a demonstration of an experimental new Java build tool "Mill" that makes use of these ideas, proving out the idea that Java build tooling has the potential to be much faster, safer, and easier than it is today.
Ram Lakshmanan
@yCrash
GC Tuning: A Masterpiece in Performance Engineering
Description: Garbage Collection Tuning is often overlooked in Enterprises. However, GC tuning dramatically reduces computing cost & drastically cuts down the response time. In this talk real case Studies from Fortune 100 companies will be shared.
In this session, you’ll gain firsthand insights into how industry leaders have approached Garbage Collection (GC) optimization to achieve significant performance improvements and save millions in infrastructure costs. We’ll analyze real GC logs, demonstrate essential tools, and reveal expert techniques used during these tuning efforts. Plus, you’ll walk away with 9 practical tips to optimize your application’s GC performance.
GC Tuning: A Masterpiece in Performance Engineering
Description: Garbage Collection Tuning is often overlooked in Enterprises. However, GC tuning dramatically reduces computing cost & drastically cuts down the response time. In this talk real case Studies from Fortune 100 companies will be shared.
In this session, you’ll gain firsthand insights into how industry leaders have approached Garbage Collection (GC) optimization to achieve significant performance improvements and save millions in infrastructure costs. We’ll analyze real GC logs, demonstrate essential tools, and reveal expert techniques used during these tuning efforts. Plus, you’ll walk away with 9 practical tips to optimize your application’s GC performance.
Merve Söylemez
@Oracle Türkiye
Java ile Bugünden Yarına: Kurumsal Öncelikler, Bulut ve Yapay Zeka
Description: Oracle Türkiye Genel Müdürü Merve Söylemez’in konuk olacağı fireside oturumumuzda, teknolojinin bugününü ve yarınını Java perspektifinden aşağıdaki konuları ele alıyoruz.
Java bugün hâlâ önemli mi?
Modern uygulamaların büyük bir kısmının temelinde yer alan Java, güçlü ekosistemi, güvenliği ve ölçeklenebilirliği ile önemini korumaya devam ediyor.
Kurumsal öncelikler: Java neden yarın da kritik olacak?
Dijital dönüşüm, sürdürülebilirlik ve operasyonel verimlilik odağında ilerleyen kurumlar için Java; güvenilir, esnek ve uzun vadeli bir yatırım olmaya devam ediyor.
Java yenilikleri: Nereye gidiyor?
Sürekli gelişen yapısı, performans iyileştirmeleri ve modern geliştirme pratiklerine uyumu ile Java, geleceğin ihtiyaçlarına hızla adapte oluyor.
Yapay zeka için Java
Makine öğrenmesi ve yapay zeka uygulamalarını destekleyen kütüphaneler ve entegrasyonlar sayesinde Java, bu alanda da güçlü bir oyuncu.
Buluta geçişi basitleştirmek
Java’nın platform bağımsızlığı ve bulut uyumlu mimarisi, kurumların buluta geçiş süreçlerini hızlandırırken maliyet ve operasyonel avantajlar sağlıyor.
Java ile Bugünden Yarına: Kurumsal Öncelikler, Bulut ve Yapay Zeka
Description: Oracle Türkiye Genel Müdürü Merve Söylemez’in konuk olacağı fireside oturumumuzda, teknolojinin bugününü ve yarınını Java perspektifinden aşağıdaki konuları ele alıyoruz.
Java bugün hâlâ önemli mi?
Modern uygulamaların büyük bir kısmının temelinde yer alan Java, güçlü ekosistemi, güvenliği ve ölçeklenebilirliği ile önemini korumaya devam ediyor.
Kurumsal öncelikler: Java neden yarın da kritik olacak?
Dijital dönüşüm, sürdürülebilirlik ve operasyonel verimlilik odağında ilerleyen kurumlar için Java; güvenilir, esnek ve uzun vadeli bir yatırım olmaya devam ediyor.
Java yenilikleri: Nereye gidiyor?
Sürekli gelişen yapısı, performans iyileştirmeleri ve modern geliştirme pratiklerine uyumu ile Java, geleceğin ihtiyaçlarına hızla adapte oluyor.
Yapay zeka için Java
Makine öğrenmesi ve yapay zeka uygulamalarını destekleyen kütüphaneler ve entegrasyonlar sayesinde Java, bu alanda da güçlü bir oyuncu.
Buluta geçişi basitleştirmek
Java’nın platform bağımsızlığı ve bulut uyumlu mimarisi, kurumların buluta geçiş süreçlerini hızlandırırken maliyet ve operasyonel avantajlar sağlıyor.
Batuhan Işıker
@Garanti BBVA
Security Test Driven Development
Description: Writing secure code reduces risks and ensures sustainable products. Early security unit tests reveal business logic vulnerabilities that conventional tools often miss, helping protect your application’s integrity before reaching production.
Secure coding practices guarantee not only the proper functioning of your software, but also the protection of critical data and business processes. In today’s threat landscape, even a minor vulnerability in the code can lead to significant security breaches and expose your entire project to risk.
The Security Test Driven Development approach emphasizes addressing security requirements from the very foundation of your code, promoting a test-focused development cycle. With security unit tests, logical flaws and vulnerabilities are systematically identified and mitigated during development enabling you to catch and resolve issues before deployment while improving code quality and organizational security posture.
In this session, we will discuss the value of secure coding for your business and software lifecycle, and demonstrate with practical examples how unit tests can help you identify vulnerabilities early. Additionally, I will present a real-world application using the Spring framework and show, step-by-step, how to write security unit tests targeting logical vulnerabilities.
Security Test Driven Development
Description: Writing secure code reduces risks and ensures sustainable products. Early security unit tests reveal business logic vulnerabilities that conventional tools often miss, helping protect your application’s integrity before reaching production.
Secure coding practices guarantee not only the proper functioning of your software, but also the protection of critical data and business processes. In today’s threat landscape, even a minor vulnerability in the code can lead to significant security breaches and expose your entire project to risk.
The Security Test Driven Development approach emphasizes addressing security requirements from the very foundation of your code, promoting a test-focused development cycle. With security unit tests, logical flaws and vulnerabilities are systematically identified and mitigated during development enabling you to catch and resolve issues before deployment while improving code quality and organizational security posture.
In this session, we will discuss the value of secure coding for your business and software lifecycle, and demonstrate with practical examples how unit tests can help you identify vulnerabilities early. Additionally, I will present a real-world application using the Spring framework and show, step-by-step, how to write security unit tests targeting logical vulnerabilities.
Rijo Sam
@ABN AMRO Bank
Spring Security, What Went Wrong and How We Fixed It
Description: Struggling with Spring Security? Stop blindly copying configs or letting AI break your app. In this talk, I reveal the most common mistakes, show live demos, and teach how to secure endpoints the right way, so you can master Spring Security with confidence!
You've added **spring-boot-starter-security** to your project. Suddenly, you have a login form and your endpoints are "secure." But when you ask your AI assistant to customize it, add a custom login, secure a new endpoint, or integrate with a database, everything breaks in mysterious ways. If you've ever felt lost configuring Spring Security or found yourself copying code without knowing why, this talk is for you.
The problem isn’t the generated code. It’s the missing mental model. In this talk, we’ll explore the most common mistakes developers make when working with Spring Security and uncover the reasons behind them. You’ll gain a clear understanding of how Spring Security actually works, so you can write and review configurations with confidence. It’s time to stop treating security as magic and start building applications you truly understand.
What you’ll take away from this talk:
- How Spring Security’s filter chain actually works
- How to set up authentication and authorization
- How to avoid and correct the most common errors
- What’s new in Spring Security 7.0
Spring Security, What Went Wrong and How We Fixed It
Description: Struggling with Spring Security? Stop blindly copying configs or letting AI break your app. In this talk, I reveal the most common mistakes, show live demos, and teach how to secure endpoints the right way, so you can master Spring Security with confidence!
You've added **spring-boot-starter-security** to your project. Suddenly, you have a login form and your endpoints are "secure." But when you ask your AI assistant to customize it, add a custom login, secure a new endpoint, or integrate with a database, everything breaks in mysterious ways. If you've ever felt lost configuring Spring Security or found yourself copying code without knowing why, this talk is for you.
The problem isn’t the generated code. It’s the missing mental model. In this talk, we’ll explore the most common mistakes developers make when working with Spring Security and uncover the reasons behind them. You’ll gain a clear understanding of how Spring Security actually works, so you can write and review configurations with confidence. It’s time to stop treating security as magic and start building applications you truly understand.
What you’ll take away from this talk:
- How Spring Security’s filter chain actually works
- How to set up authentication and authorization
- How to avoid and correct the most common errors
- What’s new in Spring Security 7.0
Hüseyin Akdoğan
@Keymate
Gateway'in Ötesinde: AI Agent'lar için Fine-Grained Authorization
Gateway'in Ötesinde: AI Agent'lar için Fine-Grained Authorization
Aboullaite Mohammed
@Spotify
Building Scalable RAG Pipelines with Java and Kubernetes
Description: Retrieval-Augmented Generation (RAG) is redefining how we build AI-powered systems, enabling dynamic, knowledge-grounded responses. But moving from prototype to production requires more than just a working model — it demands serious engineering.
You’ll learn how to:
- Structure RAG systems into microservices using Java
- Integrate vector databases (like Weaviate, Milvus, or Vespa, ...) with Java clients
- Deploy and scale RAG components efficiently in Kubernetes
- Handle retrieval timeouts, semantic cache hits, and fallback strategies
- Monitor, trace, and optimize a RAG pipeline end-to-end
This talk is geared toward developers who want to move beyond toy examples and start building scalable, production-grade AI systems.
Building Scalable RAG Pipelines with Java and Kubernetes
Description: Retrieval-Augmented Generation (RAG) is redefining how we build AI-powered systems, enabling dynamic, knowledge-grounded responses. But moving from prototype to production requires more than just a working model — it demands serious engineering.
You’ll learn how to:
- Structure RAG systems into microservices using Java
- Integrate vector databases (like Weaviate, Milvus, or Vespa, ...) with Java clients
- Deploy and scale RAG components efficiently in Kubernetes
- Handle retrieval timeouts, semantic cache hits, and fallback strategies
- Monitor, trace, and optimize a RAG pipeline end-to-end
This talk is geared toward developers who want to move beyond toy examples and start building scalable, production-grade AI systems.
Makan Sepehrifar
@Code Nomads
Capability Mesh: Enabling Orchestration through Semantic Integration
Description: This talk introduces Capability Mesh, an architectural shift from Service Mesh where services are built for composition, not consumption.I will show how MCP, human-on-the-loop, and Self-Contained Systems create a new distributed architecture designed for semantic integration in the agentic AI era.
Ten years ago, microservices chaos led to "Service Mesh" solving the discoverability problem by replacing network endpoints with meaningful references, enabling scalability and observability. It worked brilliantly for connecting services.
Today we're living through the same cycle with Generative AI and Agents, but the chaos is about integration, not networking. The old problems—discoverability, scalability, observability—are back because orchestrators now pick agents by capability, not name. Plus we face new challenges: explainability of agent actions and the possibility of unintended actions due to the non-deterministic nature of large language models.
We need the same evolutionary leap from "Service Mesh" to "Capability Mesh"—shifting from technical service contracts to semantic capability integration for autonomous AI.
## Key Topics and Takeaways:
- From OpenAPI to MCP: The Protocol Shift in Agentic AI
- MCP Gateways: The New Control Plane
- Capability Hierarchies: Composing Agents at Scale
- Human-on-the-Loop: Redefining Our Role in Agent Systems
- Policy as Code: Governing Autonomous Agents
- Observability & Safety: Tracing Agent Behavior
Capability Mesh: Enabling Orchestration through Semantic Integration
Description: This talk introduces Capability Mesh, an architectural shift from Service Mesh where services are built for composition, not consumption.I will show how MCP, human-on-the-loop, and Self-Contained Systems create a new distributed architecture designed for semantic integration in the agentic AI era.
Ten years ago, microservices chaos led to "Service Mesh" solving the discoverability problem by replacing network endpoints with meaningful references, enabling scalability and observability. It worked brilliantly for connecting services.
Today we're living through the same cycle with Generative AI and Agents, but the chaos is about integration, not networking. The old problems—discoverability, scalability, observability—are back because orchestrators now pick agents by capability, not name. Plus we face new challenges: explainability of agent actions and the possibility of unintended actions due to the non-deterministic nature of large language models.
We need the same evolutionary leap from "Service Mesh" to "Capability Mesh"—shifting from technical service contracts to semantic capability integration for autonomous AI.
## Key Topics and Takeaways:
- From OpenAPI to MCP: The Protocol Shift in Agentic AI
- MCP Gateways: The New Control Plane
- Capability Hierarchies: Composing Agents at Scale
- Human-on-the-Loop: Redefining Our Role in Agent Systems
- Policy as Code: Governing Autonomous Agents
- Observability & Safety: Tracing Agent Behavior
Çiğdem Şengül
@Himera Agile Consulting
Çekirdek IT Ekipleri vs. Yeni Nesil Yapay Zeka Ekipleri
Çekirdek IT Ekipleri vs. Yeni Nesil Yapay Zeka Ekipleri
Doğa Güçlü Deniz
@Himera Agile Consulting
Çekirdek IT Ekipleri vs. Yeni Nesil Yapay Zeka Ekipleri
Çekirdek IT Ekipleri vs. Yeni Nesil Yapay Zeka Ekipleri
Ümit Köse
@Türksat
Modern Java ile Design Patterns’in Fonksiyonel Evrimi
Description: Java her güncellemeyle değişiyorken, bakış açımızı değiştiriyor muyuz? Design patterns'in Java 25 ile evrimini canlı örneklerle keşfedin. Lambda'lar, pattern matching, sealed classes - aynı problemler, %80 daha az kod. OOP meets Functional, gerçek refactoring örnekleriyle. Hem klasik hem modern!
Bu konuşma, klasik Gang of Four design pattern'lerinin modern Java (17–25) ile nasıl daha sade, daha güçlü ve daha fonksiyonel bir hâle geldiğini pratik örneklerle göstermeyi amaçlıyor. Design patterns hâlâ güncel ve değerli; ancak Java’nın lambda ifadeleri, functional interfaces, Stream API, sealed classes ve pattern matching gibi modern özellikleriyle birlikte bu pattern'lerin uygulanma biçimleri dramatik şekilde değişti. Bu dönüşüm, hem boilerplate’i azaltıyor hem de okunabilirlik ve test edilebilirlik açısından gerçek kazanımlar sağlıyor.
25 dakikalık bu oturumda, Strategy, Builder, Template Method, Decorator ve Visitor gibi yaygın kullanılan pattern'lerin klasik hâlleriyle modern fonksiyonel versiyonlarını canlı kodlama ile yan yana göreceksiniz. Örneğin Strategy Pattern, interface + concrete sınıf yapısından yalnızca birkaç satırlık bir lambda ifadesine dönüşürken; Decorator Pattern fonksiyon kompozisyonu ve andThen() zincirleriyle çok daha temiz bir forma kavuşuyor. Aynı şekilde Template Method, abstract sınıf hiyerarşilerinden sıyrılıp higher-order function yaklaşımıyla daha esnek bir mimariye evriliyor. Java 25’in sealed classes, record patterns ve gelişmiş pattern matching özellikleri ise Visitor Pattern gibi karmaşık yapıları son derece sade bir hâle getiriyor.
Bu konuşmanın odak noktası yalnızca “nasıl uygulanır” değil; aynı zamanda “ne zaman OOP, ne zaman fonksiyonel yaklaşım tercih edilmeli?” sorusuna gerçek dünya deneyimleriyle yanıt vermek olacak. Kısa performans notları, okunabilirlik-kısalık dengesi, ekip içi kod review etkileri ve legacy kodu modern Java’ya taşıma stratejileri de ele alınacak.
Canlı kodlama formatında ilerleyecek oturum, özellikle Java 8+ ile çalışan, design pattern’lere aşina olan ve modern Java özelliklerini gerçek projelerde nasıl uygulayabileceğini merak eden geliştiriciler için ideal. Katılımcılar, konuşma sonunda birkaç pattern’i kendi projelerinde hemen modernize edebilecek bilgi ve örneklere sahip olacak.
Modern Java ile Design Patterns’in Fonksiyonel Evrimi
Description: Java her güncellemeyle değişiyorken, bakış açımızı değiştiriyor muyuz? Design patterns'in Java 25 ile evrimini canlı örneklerle keşfedin. Lambda'lar, pattern matching, sealed classes - aynı problemler, %80 daha az kod. OOP meets Functional, gerçek refactoring örnekleriyle. Hem klasik hem modern!
Bu konuşma, klasik Gang of Four design pattern'lerinin modern Java (17–25) ile nasıl daha sade, daha güçlü ve daha fonksiyonel bir hâle geldiğini pratik örneklerle göstermeyi amaçlıyor. Design patterns hâlâ güncel ve değerli; ancak Java’nın lambda ifadeleri, functional interfaces, Stream API, sealed classes ve pattern matching gibi modern özellikleriyle birlikte bu pattern'lerin uygulanma biçimleri dramatik şekilde değişti. Bu dönüşüm, hem boilerplate’i azaltıyor hem de okunabilirlik ve test edilebilirlik açısından gerçek kazanımlar sağlıyor.
25 dakikalık bu oturumda, Strategy, Builder, Template Method, Decorator ve Visitor gibi yaygın kullanılan pattern'lerin klasik hâlleriyle modern fonksiyonel versiyonlarını canlı kodlama ile yan yana göreceksiniz. Örneğin Strategy Pattern, interface + concrete sınıf yapısından yalnızca birkaç satırlık bir lambda ifadesine dönüşürken; Decorator Pattern fonksiyon kompozisyonu ve andThen() zincirleriyle çok daha temiz bir forma kavuşuyor. Aynı şekilde Template Method, abstract sınıf hiyerarşilerinden sıyrılıp higher-order function yaklaşımıyla daha esnek bir mimariye evriliyor. Java 25’in sealed classes, record patterns ve gelişmiş pattern matching özellikleri ise Visitor Pattern gibi karmaşık yapıları son derece sade bir hâle getiriyor.
Bu konuşmanın odak noktası yalnızca “nasıl uygulanır” değil; aynı zamanda “ne zaman OOP, ne zaman fonksiyonel yaklaşım tercih edilmeli?” sorusuna gerçek dünya deneyimleriyle yanıt vermek olacak. Kısa performans notları, okunabilirlik-kısalık dengesi, ekip içi kod review etkileri ve legacy kodu modern Java’ya taşıma stratejileri de ele alınacak.
Canlı kodlama formatında ilerleyecek oturum, özellikle Java 8+ ile çalışan, design pattern’lere aşina olan ve modern Java özelliklerini gerçek projelerde nasıl uygulayabileceğini merak eden geliştiriciler için ideal. Katılımcılar, konuşma sonunda birkaç pattern’i kendi projelerinde hemen modernize edebilecek bilgi ve örneklere sahip olacak.
İbrahim Atay
@Citibank
LLM-Powered Unit Testing: Writing Reliable Java Tests with AI Assistance
Description: Discover how Large Language Models (LLMs) like OpenAI and Copilot can supercharge your Java unit testing workflow. In this talk, you'll learn to generate, validate, and scale test cases using AI, boosting coverage while saving time.
Unit testing is an essential part of modern software development, yet it remains one of the most time-consuming and often-neglected practices, especially under pressure to deliver. With the rise of Large Language Models (LLMs), developers now have powerful assistants capable of generating boilerplate code, documentation, and increasingly… test cases.
In this talk, we’ll explore how LLMs like OpenAI, Claude, and GitHub Copilot can assist Java developers in writing, optimizing, and scaling unit tests. We’ll cover the strengths and limitations of AI-generated tests, how to evaluate their quality, and what architectural patterns enable safe and effective LLM integration.
Through practical Java code examples and live demonstrations, you'll learn how to:
- Generate meaningful test cases for real-world Java methods
- Use LLMs responsibly by designing effective prompts and validation layers
- Automate unit test suggestions in your CI pipeline or IDE
- Combine LLM support with code quality tools like SonarQube and JaCoCo
This session will be valuable for any Java developer or architect seeking to enhance test coverage, minimize repetitive effort, and leverage AI-enhanced development workflows without compromising quality.
LLM-Powered Unit Testing: Writing Reliable Java Tests with AI Assistance
Description: Discover how Large Language Models (LLMs) like OpenAI and Copilot can supercharge your Java unit testing workflow. In this talk, you'll learn to generate, validate, and scale test cases using AI, boosting coverage while saving time.
Unit testing is an essential part of modern software development, yet it remains one of the most time-consuming and often-neglected practices, especially under pressure to deliver. With the rise of Large Language Models (LLMs), developers now have powerful assistants capable of generating boilerplate code, documentation, and increasingly… test cases.
In this talk, we’ll explore how LLMs like OpenAI, Claude, and GitHub Copilot can assist Java developers in writing, optimizing, and scaling unit tests. We’ll cover the strengths and limitations of AI-generated tests, how to evaluate their quality, and what architectural patterns enable safe and effective LLM integration.
Through practical Java code examples and live demonstrations, you'll learn how to:
- Generate meaningful test cases for real-world Java methods
- Use LLMs responsibly by designing effective prompts and validation layers
- Automate unit test suggestions in your CI pipeline or IDE
- Combine LLM support with code quality tools like SonarQube and JaCoCo
This session will be valuable for any Java developer or architect seeking to enhance test coverage, minimize repetitive effort, and leverage AI-enhanced development workflows without compromising quality.
Hamza Metin Yarcı
@Mavidev
Bankacılıkta Bugün İçin Kodla, Yarın İçin Tasarla
Bankacılıkta Bugün İçin Kodla, Yarın İçin Tasarla
Anton Arhipov
@JetBrains
Spring Debugger
Description: Spring Boot feels like magic until it doesn’t. When your app misbehaves, the Spring Debugger in IntelliJ IDEA reveals what’s really happening: property origins, bean wiring, transactions, DB connections, and more right in your IDE. Debug faster, learn deeper, and make Spring an open book.
Spring is known for its flexibility and power—but also for the invisible wiring that can make understanding what’s really happening in your application tricky. Beans appear as if by magic, properties seem to resolve themselves, and sometimes it’s hard to see how everything fits together.
In this session, we’ll peel back the layers of Spring’s runtime behavior and learn how to debug Spring applications with precision. We’ll explore how to trace bean creation and injection, follow property resolution across profiles and configuration sources, and inspect transactions and data connections—all while keeping your mental model of the application crystal clear.
We’ll use the Spring Debugger in IntelliJ IDEA to make these insights visible, giving you the ability to explore the framework’s internals interactively instead of guessing or grepping through logs.
You’ll learn how to:
* Understand how Spring resolves configuration and property values.
* See which beans are created, injected, and why specific candidates are chosen.
* Track transactions, data sources, and runtime context in real time.
* Turn Spring’s dynamic behavior into something you can explore, reason about, and explain.
By the end, you’ll not only feel more confident navigating Spring Boot’s inner workings—you’ll see how debugging can become a tool for learning and mastery, not just fixing what’s broken.
Spring Debugger
Description: Spring Boot feels like magic until it doesn’t. When your app misbehaves, the Spring Debugger in IntelliJ IDEA reveals what’s really happening: property origins, bean wiring, transactions, DB connections, and more right in your IDE. Debug faster, learn deeper, and make Spring an open book.
Spring is known for its flexibility and power—but also for the invisible wiring that can make understanding what’s really happening in your application tricky. Beans appear as if by magic, properties seem to resolve themselves, and sometimes it’s hard to see how everything fits together.
In this session, we’ll peel back the layers of Spring’s runtime behavior and learn how to debug Spring applications with precision. We’ll explore how to trace bean creation and injection, follow property resolution across profiles and configuration sources, and inspect transactions and data connections—all while keeping your mental model of the application crystal clear.
We’ll use the Spring Debugger in IntelliJ IDEA to make these insights visible, giving you the ability to explore the framework’s internals interactively instead of guessing or grepping through logs.
You’ll learn how to:
* Understand how Spring resolves configuration and property values.
* See which beans are created, injected, and why specific candidates are chosen.
* Track transactions, data sources, and runtime context in real time.
* Turn Spring’s dynamic behavior into something you can explore, reason about, and explain.
By the end, you’ll not only feel more confident navigating Spring Boot’s inner workings—you’ll see how debugging can become a tool for learning and mastery, not just fixing what’s broken.
Haoyi Li
Scripting on the JVM with Java, Scala, and Kotlin
Description: This talk will explore use of JVM languages as scripting languages, replacing Python or Bash. This is a capability that is not often used by Java developers traditionally focused on backend systems, and this talk will explore its pros and cons and demonstrate what is needed to do this successfully.
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.
Description: This talk will explore use of JVM languages as scripting languages, replacing Python or Bash. This is a capability that is not often used by Java developers traditionally focused on backend systems, and this talk will explore its pros and cons and demonstrate what is needed to do this successfully.
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.
Halis Yılboğa
@TÜBİTAK BİLGEM BTE
Building AI powered search in Java with Apache Solr
Description: Supercharge your Java search with Solr's cutting-edge features!
Semantic Search: Go beyond keywords with word embeddings and vector search. Delight users with "More Like This" suggestions powered by AI.
Intelligent Product Matching: Boost e-commerce with AI-driven product matching.
Is your search engine still just a keyword matcher? It's time to level up. This session will show you how to supercharge your Java applications with Apache Solr's cutting-edge AI capabilities.
We'll dive into the world of semantic search, moving beyond simple text matching to understand user intent. You will learn how to:
Implement Vector Search: Harness the power of word embeddings to find conceptually similar results, not just textually identical ones.
Delight Users: Create "More Like This" features that are genuinely helpful and intuitive.
Boost E-commerce: Build an AI-driven product matching system that understands product attributes and dramatically improves conversions.
Join us to transform your search from a simple utility into your application's most powerful, intelligent feature.
Building AI powered search in Java with Apache Solr
Description: Supercharge your Java search with Solr's cutting-edge features!
Semantic Search: Go beyond keywords with word embeddings and vector search. Delight users with "More Like This" suggestions powered by AI.
Intelligent Product Matching: Boost e-commerce with AI-driven product matching.
Is your search engine still just a keyword matcher? It's time to level up. This session will show you how to supercharge your Java applications with Apache Solr's cutting-edge AI capabilities.
We'll dive into the world of semantic search, moving beyond simple text matching to understand user intent. You will learn how to:
Implement Vector Search: Harness the power of word embeddings to find conceptually similar results, not just textually identical ones.
Delight Users: Create "More Like This" features that are genuinely helpful and intuitive.
Boost E-commerce: Build an AI-driven product matching system that understands product attributes and dramatically improves conversions.
Join us to transform your search from a simple utility into your application's most powerful, intelligent feature.
Empowering Developers in the Age of AI
IMPRESSIONS
SPONSORS
Java Day Istanbul supported by industry sponsors. We appreciate their commitment to the Istanbul Java User Group and we greatly value their active contribution to the JAVA software and technology community. Please fill the form below and we will get back to you with our sponsorship packages with our services and our conditions.
Sponsors
Training Partner
Media Partner
Strategic Partners
Community Supporters
Meet Our Team
Anshika Koul
Senior Software Engineer
@Posten Bring AS
Altuğ B. Altıntaş
JUG Leader, Enterprise Architect @ING Bank
Akın Kaldıroğlu
Chief Learning Officer @BackendGuru
Nursel Cıbır
Business Agility Consultant
@Agile Kanban Istanbul
Jonathan Vila
Staff Developer Advocate
@Sonar
Venue
Venue
ABOUT THIS VENUE:
Elite World Grand İstanbul Küçükyalı in Maltepe, one of the best locations on Istanbul’s Asian side, you can soak in the perfect view of the sea and the Princes’ Islands!
FULL ADDRESS:
İnönü Caddesi No: 22 Küçükyalı
Maltepe - Istanbul / Turkey
Contact
bottom of page