Cheaper or Better? The Strategic Choice Behind AI Adoption
Miriam SasseTwo AI adoption strategies: efficiency gains or customer value. Incentives, metrics and traps (automating a broken process).
▶ WatchTalks I didn't see
I could not be everywhere. These are the talks I had selected but missed, grouped by theme — from the closest to AI adoption in the enterprise to general Java. Most of them have a recording: grab a coffee ☕
AI adoption & enablement in the enterprise
Section titled “AI adoption & enablement in the enterprise”Field reports, strategy, measurement and organisation: the core of the AI enablement lead role.
0 to 350 Engineers on Claude Code: What Could Go Wrong?
Nicolas MartignoleBack Market rolls out Claude Code to 350 engineers: PMs who code, costs, 50-file PRs and guardrails. The closest thing to a real adoption plan.
▶ WatchAI Made Building Cheap. Now Decisions Are the Bottleneck.
Dave Sammut, João RosaAI makes delivery fast, so decisions become the bottleneck. How to turn architecture discussions into maps, risks and options.
▶ WatchCoding-agent workflows
Section titled “Coding-agent workflows”Specs, loops, harness, parallel agents, tooling: how to work with a coding agent.
Can AI Build a Jakarta EE Runtime? The Vidocq Experiment
Antoine Sabot-Durand, yann Blazart, Sebastien BlancA Jakarta EE / MicroProfile runtime built with an agent constrained by specs and TCKs: what worked, what failed, the workflow kept.
▶ WatchBeyond Code: The Intent-Driven IDE for the Agentic Age
Bruno BorgesIntent-driven IDE: from spec to PR with agents in parallel workspaces, live demo and workflow extension.
▶ WatchTeaching a New Dog Old Tricks: Give the Agent a Debugger
Anton ArhipovGive the IntelliJ debugger to the agent (logpoints, breakpoints) instead of letting it add println.
▶ WatchFrom Developer to Story Builder: Engineering a Novel, an ARG, and a Transmedia World with AI
Raphael SemeteysCo-writing a novel and an ARG with AI using dev practices: context engineering, spec-driven, separated roles, anti-drift governance.
▶ Watch223 Pull Requests in 11 Days: The Recipe for Building Real Software with AI Agents
Julien Dubois83,000 lines in 11 days without opening the IDE: specs, test harness, parallel agents, overnight builds, merge in the morning. The most concrete case study.
▶ WatchState of Agentic Coding
Daniela PetruzalekOverview of recent advances (MCP, skills…), demos on everyday dev work, how to build trust in the agent, tour of Google Antigravity.
▶ WatchIs the IDE dead ?
Martin LippertIDE or terminal: what does the developer still need when the agent becomes the power user of the tools?
▶ WatchHardwood: Building a Parquet Parser From Scratch (With a Little Help From AI)
Gunnar MorlingJava Parquet parser written with Claude Code as a companion: mostly a performance talk, with an experience report on demanding code.
▶ WatchWe created our own AI orchestrator and it became so much more!
Ricky van RijnHome-grown agent orchestrator (Embabel, skills, specs) with a dashboard, and how it changes developer / analyst roles.
No video yetControl, quality & guardrails for generated code
Section titled “Control, quality & guardrails for generated code”Staying in charge when the agent codes: deterministic rules, review, measurement.
Testing Architecture Made Simple
Gunter RotsaertArchUnit and Taikai turn architecture decisions into automated tests: the deterministic guardrail against agents.
▶ WatchAgents With Seatbelts: Practical Ways to Keep AI Code Gen Under Control
Jonathan Vila LópezMCP and agent best practices, code stability across models, how to measure the real gain from AI.
No video yetSoftware engineering in the age of AI agents: who’s in control?
Bert Jan SchrijverConcrete stories (prod database incident, app generated from a spec in 45 min) and a framework: experiment, deploy carefully, raise the testing bar.
▶ WatchOwning the Code You Didn't Type
Renato Guerra CavalcantiWhen AI writes faster than we can read: phased delivery, guided walk-throughs, human-centered review.
▶ WatchIntroducing Agent Control Theory: The System Runs the Model
Mark PollackTreat coding agents as measurable systems: log every run, judge against specs, optimize prompts / tools / models.
▶ Watch10 boring reasons why Java is the best ecosystem for agentic development
Ed BurnsJava's "boring" strengths (types, tests, static analysis, builds) that limit agent drift, with an agent-safe checklist.
No video yetAgent security & governance
Section titled “Agent security & governance”MCP / skills risks, human-in-the-loop, AI Act, identity and platform.
AI Security Deep Dive, Securing Agents and AI Applications
Brian VermeerSecurity for coding agents and AI apps: prompt injection, tool abuse, poisoned memory, defenses and tooling, live demos.
▶ WatchDon’t burn your fingers in the heat of the AI (r)evolution. (And get an OpenClaw).
Maarten VandeperreMaking agentic AI enterprise-ready: security, governance, observability and sandboxing handled by the platform.
▶ WatchThe Importance of HITL to Avoid Chaos and EU Regulations for AI
Juan Antonio BreñaHuman-in-the-loop for coding agents in the enterprise: approval gates, audit, AI Act, without losing CI/CD speed.
▶ WatchBuilding an Agent Ready Platform
Timo SalmPlatform to take agents to production: secure model access, agent identity, inherited guardrails, cost tracking.
▶ WatchToxic Flows in MCP Servers and Skills: What Your Agent Blindly Trusts
Brian Vermeer"Toxic flows": trust paths in MCP servers and skills that leak context or trigger unintended actions. Watch before sharing skills internally.
▶ WatchSay Goodbye to LLM API keys with Workload Identity Federation
Siarhei BiarozkinReplace LLM API keys with short-lived tokens via Workload Identity Federation (SPIFFE, Kubernetes).
▶ WatchWho to blame? The AI, The Programmer, or The Prompt?
Makan SepehrifarInteractive trial: who is to blame (AI, programmer, prompt) for five real CVEs in AI tools.
▶ WatchLegacy & modernization
Section titled “Legacy & modernization”Migrating or exposing existing systems, with or without AI.
No Country For Old APIs. Wrapping legacy Java services as MCP tools without rewriting them.
Željko KozinaExpose legacy Java (REST, SOAP) to agents via MCP without rewriting: auth, tool sprawl, dangerous retries.
▶ WatchEngineering Lessons from Working on Software Older Than Me
Guust YsebieLessons from maintaining very old PDF technology: compatibility, performance, security, testing.
▶ WatchModernizing legacy code with AI: hands-on examples & lessons learned
Koen Vanderkimpen, Wim LambrechtSmals Research: Java EE to Jakarta EE (and COBOL) migration with AI. Where it speeds up, where it fails, how to stay verifiable.
▶ WatchStrangling the dragon: modernizing legacy systems with the Strangler Fig pattern
Daniel RanelandModernize legacy step by step with the Strangler Fig and DDD, no big bang or full rewrite.
No video yetArchitecture, debt & documentation
Section titled “Architecture, debt & documentation”Architecture, design, docs and APIs in the age of agents.
Beyond system.out.println: Applying a Debugging Mindset at Scale
Cees BosDebugging intermittent failures in distributed systems: structured observability, OpenTelemetry, Grafana.
No video yetAsyncAPI in action: OpenAPI for event-driven applications
Michel SchudelAsyncAPI: one spec generates Kafka producer, consumer, DTOs and docs. Same contract-first logic as spec-driven.
▶ WatchWrap, Reshape, or Redesign: Retrofitting Your APIs for a World of Agents
Horacio GonzalezAdapting human-oriented APIs to agents that take them literally: wrap, reshape or redesign.
▶ WatchWhy Documentation Won’t Write Itself—and AI Won’t Save You
Christian HeitzmannAI won't write your docs for you: docs close to code, separating architecture / API / code docs, preventing decay.
▶ WatchHow to Architect Green Software
Ioannis KolaxisGreen software: weighing carbon footprint against maintainability, availability and performance.
▶ WatchYou thought technical debt was bad? Meet design debt!
Julien Topçu, Josian ChevalierDesign debt, deeper than technical debt: how it forms, how to detect and avoid it.
No video yetSoftware Monuments - Architectural Lessons from the Systems That Shaped Computing
Tobias VossArchitecture lessons from the systems that shaped computing.
No video yetCulture, teams & community
Section titled “Culture, teams & community”The human side: coaching, conflicts, self-doubt, open source, DevEx.
Emergent DevEx: The Good, the Bad, and the Legacy
Makan Sepehrifar, Soroosh KhodamiDevEx degrades as organisations grow; why copying Google or Netflix is a trap.
▶ WatchSurviving the "AI Slopageddon": How AI shakes the foundations of Open Source
Raphael SemeteysAI-generated code overwhelms open-source maintainers (sloppy PRs, fake bug reports); ideas for defense.
▶ Watchrm -rf self-doubt: Debugging Impostor Syndrome for Developers
Vincent MayersNeuroscience of impostor syndrome, handling self-doubt in code review and design discussions.
▶ WatchHelp! I hate my team mate!
Jessica SiewertTeam conflicts when working styles clash: experience reports and options.
▶ WatchBuilding application agents
Section titled “Building application agents”Building agents inside apps (Spring AI, LangChain4j, Embabel, Docker Agent…): excluded from the selection criterion, kept for reference.
Out of my scope — show the 11 talks
From AI to Agent: A Field Guide to Agentic Patterns in LangChain4j
Loïc MagnetteLangChain4j agentic patterns: sequential workflows, loops, supervisors, human checkpoints.
▶ WatchIntroduction to Spring AI
Timo SalmIntroduction to Spring AI: chat, RAG, tool calling, observability, MCP. Prepares for the afternoon Spring lab.
▶ WatchBuilding Autonomous AI Agents with Spring
Timo Salm, Sandra AhlgrimmWorkshop: autonomous agents with Spring AI and MCP (memory, RAG, tools, observability, security).
No video yetBuilding AI Agents from First Principles
Adib SaikaliAn agent built from scratch with Spring AI: plan-act-observe loop, tools, memory, evaluation.
▶ WatchSpecialized AI Agents: from concept to production
David GageotCheap specialized agents with Docker Agent: creation, evaluation, cost monitoring.
▶ WatchBuild the Builder: A Java Coding Agent in Your IDE with ACP
Vadim BriliantovBuild a Java/Kotlin coding agent plugged into IntelliJ via ACP, and orchestrate Codex, Claude, Gemini.
▶ WatchEval Driven Development: Why you need it to make your Agentic Application survive Production
Urs PeterEval-Driven Development: layered evals and golden datasets for agentic apps in production.
▶ WatchA Year of Agentic AI Evolution: Lessons Learned Building Production-grade Agentic Systems
Mario Fusco, Kevin DuboisA year of the LangChain4j agentic framework: patterns, orchestration, production lessons.
▶ WatchCreate your own AI Agents without code
David GageotDocker Agent: no-code agents, benchmarking and multi-agent workflows, with any LLM.
▶ WatchChoose your own adventure in agentic design patterns
Guillaume LaforgeAdvanced multi-agent patterns: routing, swarms, planning, LLM-as-Judge, agent loops.
▶ WatchEmbabel Agent v1.5
Rod JohnsonEmbabel: JVM agent framework with deterministic planning (GOAP) on Spring AI.
No video yetLocal models & inference
Section titled “Local models & inference”Running models locally or in pure Java (GDPR, sensitive data).
Run Frontier Open Models on consumer hardware
Prince CanumaFrontier open models on Mac with MLX: unified memory, batching, KV cache, benchmarks.
▶ WatchQuixotic: Pure Java Inference at Native Speed
Alfonso² PeterssenAI inference in pure Java, no Python or native code: sensitive data stays inside the JVM (regulated sectors).
▶ WatchGeneral Java & tooling
Section titled “General Java & tooling”Non-AI: language, frameworks, tools.
Learning modern Java the playful way
Marit van Dijk, Piotr PrzybylPlayful modern Java: Structured Concurrency, Stream Gatherers.
▶ Watch🧰 Dev Containers: the ultimate toolbox for developers?
Stéphane PhilippartDev containers: dev environment as code, a base for isolating and reproducing a coding agent.
No video yetStructured Concurrency Looms
Alan Bateman, Viktor KlangStructured Concurrency (Loom): simpler cancellation, error handling, observability.
▶ WatchBootiful Spring Boot 4
Josh LongSpring Boot 4 on Java 25+: modularity, virtual threads, Leyden, GraalVM, AI.
▶ WatchSnake in 10 Lines: Learning More by Coding Less
Guus de WitSnake in 10 lines of plain JDK: code less to understand more.
▶ WatchWhen ORM Becomes OMG: Performance Pitfalls in JPA and Friends
Jos RoseboomJPA performance pitfalls: unexpected loading, transactions, connections.
▶ Watch"You May Now Optimize the Wedding": Optimizing the Seating, Bachelor Party, and Honeymoon with Constraint AI
Geoffrey De Smet, Tom CoolsConstraint optimization with Timefold, illustrated by a wedding seating plan.
▶ WatchConquest, time travel and explosions using Git
Alexander ChatzizachariasLearn Git (branches, rebase, detached head) through a real-time strategy game.
▶ WatchBootiful Security
Daniel Garnier-Moiroux, Josh LongPasskeys with Spring Security 7 to replace passwords.
▶ WatchAll you need to know about Maven 4!
Matthias BüngerMaven 4: why, and the key new features.
No video yet