docs(cfp): Adding proposals for Devoxx Be 2026 (#827)

jabrena · web-flow · commit 30fcf13cef40 · 2026-06-09T15:27:37.000+02:00
diff --git a/documentation/conferences/cfp/dvbe26/technical-ai-workshop/ABSTRACT_TALK.md b/documentation/conferences/cfp/dvbe26/technical-ai-workshop/ABSTRACT_TALK.md
@@ -0,0 +1,20 @@
+Abstract
+
+**Technical Workshop on AI-Native Tooling for Java Development**
+
+For nearly three decades, enterprise Java software has been built manually by software engineers. In 2023, a new generation of AI-powered tools emerged — and tools like Cursor AI, Claude Code, and GitHub Copilot are now fundamentally changing how software is designed, implemented, and operated across the full SDLC.
+This workshop introduces an opinionated AI-native workflow for evolving modern Java Enterprise SDLC practices through reusable Skills, Agents, Commands, and MCP Servers — structured to answer the key engineering questions every team faces: what to build, why to build it, and how to build it well. githubusercontent
+The workshop is designed for all levels and covers the following topics in a practical, hands-on way:
+
+- How the SDLC has evolved with AI-native engineering workflows
+- The three core workflows: Prompt Engineering, Agent-driven Engineering, and Pipelines
+- How to use Commands to automate repeatable actions across the development lifecycle — from /update-issue-description and /implement to /verify
+- How to compose reusable Skills across analysis, design, implementation, and operations phases
+- How to configure and delegate to Agents such as @robot-coordinator, @robot-java-coder, and framework-specific agents for Spring Boot, Quarkus, and Micronaut
+- How to extend agent capabilities with MCP Servers for JavaDocs, JDBC, and Grafana
+- How to write a solid AGENTS.md and structure specs for safe, effective agent delegation
+- Practical limitations to understand: non-determinism, model variability, and how to mitigate them with clear goals and validation checkpoints
+
+The demos will walk through real examples using Spring Boot, Quarkus, and Micronaut — showing how Commands, Skills, and Agents address engineering challenges across classic areas such as Maven build systems, object-oriented design, secure coding, performance testing with JMeter, and profiling with Async Profiler and OpenJDK tools.
+
+The workshop will be delivered in English.
diff --git a/documentation/conferences/cfp/dvbe26/technical-ai-workshop/ELEVATOR-PITCH.md b/documentation/conferences/cfp/dvbe26/technical-ai-workshop/ELEVATOR-PITCH.md
@@ -0,0 +1,7 @@
+Elevator pitch
+
+Elevator pitch in English
+
+The Java SDLC is being rewritten. Commands, Skills, Agents, and MCP Servers are turning what used to be manual, repetitive engineering work into structured, AI-native workflows — and teams that adopt them early are shipping faster with fewer mistakes.
+
+In just 90 minutes, attendees will get hands-on with an opinionated AI-native workflow built specifically for Java enterprise development — learning how to compose reusable Skills, delegate implementation to framework-specific Agents, and wire up MCP Servers across the full lifecycle, from planning and architecture all the way through profiling and operations — using Spring Boot, Quarkus, and Micronaut as the playground.
diff --git a/documentation/conferences/cfp/dvbe26/the-importance-of-hitl-to-avoid-chaos/ABSTRACT_TALK.md b/documentation/conferences/cfp/dvbe26/the-importance-of-hitl-to-avoid-chaos/ABSTRACT_TALK.md
@@ -0,0 +1,14 @@
+Abstract
+
+**The Importance of HITL to Avoid Chaos**
+
+AI coding agents are rapidly becoming first-class citizens in enterprise software workflows. But as teams delegate more decisions to autonomous agents — generating migrations, refactoring service boundaries, or modifying database schemas — a critical question emerges: who is responsible when something goes wrong?
+This talk explores the discipline of Human-in-the-Loop (HITL) design, and how to build AI-assisted workflows that keep humans meaningfully in control without sacrificing the productivity gains that make agents worth adopting in the first place.
+The talk will cover in detail:
+
+- What HITL means in practice: approval gates, confidence thresholds, and audit trails
+- High-stakes scenarios in microservices architectures: agent-driven database migrations, schema drift across service boundaries, and cascading failures triggered by automated refactoring
+- How the EU AI Act classifies autonomous coding agents and what compliance obligations teams must consider when deploying them in enterprise environments
+- Practical patterns for embedding HITL checkpoints into CI/CD pipelines without turning every deployment into a bottleneck
+
+The session will demonstrate these concepts through real-world scenarios where the absence of human oversight led to cascading failures — and show how the right HITL design would have caught them early — giving teams a framework to move fast with AI agents without losing control of their systems.
diff --git a/documentation/conferences/cfp/dvbe26/the-importance-of-hitl-to-avoid-chaos/ELEVATOR-PITCH.md b/documentation/conferences/cfp/dvbe26/the-importance-of-hitl-to-avoid-chaos/ELEVATOR-PITCH.md
@@ -0,0 +1,7 @@
+Elevator pitch
+
+Elevator pitch in English
+
+AI coding agents are powerful — but an agent with full autonomy over your microservices and databases is also an incident waiting to happen. Schema migrations applied without review, cascading refactors across service boundaries, and automated deployments with no approval gate are real risks teams are already facing.
+
+In just 30 minutes, attendees will learn why Human-in-the-Loop is not a limitation on AI productivity but a safety architecture — and how the EU AI Act is making it a legal requirement for many enterprise systems — walking away with practical patterns to keep humans meaningfully in control without slowing teams down.
diff --git a/documentation/conferences/cfp/dvbe26/validating-skills-in-your-ci/ABSTRACT_TALK.md b/documentation/conferences/cfp/dvbe26/validating-skills-in-your-ci/ABSTRACT_TALK.md
@@ -0,0 +1,14 @@
+Abstract
+
+**Validationg Skills in your CI**
+
+AI coding agents are rapidly becoming first-class citizens in enterprise software workflows. But as teams scale their use of agent skills — structured instructions that guide AI behavior across tasks like code generation, testing, and refactoring — a critical question emerges: how do you know your skills actually work as intended?
+This talk explores the discipline of agent skill validation, and how to build robust CI pipelines that treat skills as first-class artifacts worthy of the same rigor applied to production code.
+
+The talk will cover in detail:
+
+- What skill validation means and why it matters: correctness, safety, and behavioral consistency
+- A multi-tool validation strategy combining static analysis with behavioral inspection
+- How to integrate skill validation into GitHub Actions pipelines alongside conventional pre-commit hooks and commit message linting
+
+The session will demonstrate these concepts through a real-world CI pipeline that validates agent skills across multiple dimensions — from schema conformance to behavioral policy enforcement — giving teams the confidence to ship AI-assisted workflows at enterprise scale.
diff --git a/documentation/conferences/cfp/dvbe26/validating-skills-in-your-ci/ELEVATOR-PITCH.md b/documentation/conferences/cfp/dvbe26/validating-skills-in-your-ci/ELEVATOR-PITCH.md
@@ -0,0 +1,7 @@
+Elevator pitch
+
+Elevator pitch in English
+
+AI agent skills aren't just configuration files — they're the backbone of how your team's coding agents think and behave. Yet most teams ship them with zero validation, one bad skill away from broken workflows at scale.
+
+In just 30 minutes, attendees will learn how to treat skills as production-grade artifacts, building a CI pipeline that validates them across multiple dimensions — from schema conformance to behavioral policy enforcement — and walk away with the confidence to ship AI-assisted workflows in enterprise environments.
