Workshops: Agentic Coding by Tech Stack

Agentic Coding with Java & Spring Boot: Enterprise Workshop

Java enterprise codebases are large, layered and slow to build, a tough environment for coding agents that learn by running commands. This workshop shows Spring Boot teams how to give Claude Code, GitHub Copilot, JetBrains Junie or OpenAI Codex fast feedback loops, the right slice of context and clear architecture rules, so agents deliver reliable changes in real projects, including framework and JDK upgrades.

Who is this workshop for?

  • Java developers in Spring Boot and Jakarta EE projects
  • Software architects responsible for large multi-module systems
  • Teams planning upgrades of Java versions or Spring Boot major versions
  • IntelliJ IDEA, Eclipse or VS Code users who want to add terminal agents
  • Tech leads in banking, insurance, industry and public sector IT with long-lived Java systems

What you will take away

  • Make Maven or Gradle builds agent-friendly with targeted commands for single modules and single tests.
  • Have agents implement features across controller, service and repository layers that respect your architecture rules.
  • Generate unit, slice and integration tests with JUnit, Mockito and Testcontainers that actually catch regressions.
  • Plan Spring Boot and JDK upgrades in which OpenRewrite recipes and agents work hand in hand.
  • Combine IDE assistants and terminal agents sensibly in a team working with JetBrains IDEs or Eclipse.

Agenda

  1. Module 1

    Enterprise Java meets agents

    • Why build time and project size shape agent workflows
    • Claude Code, OpenAI Codex, GitHub Copilot and JetBrains AI Assistant with Junie compared
    • Exercise: let the agent document a module's dependencies and entry points
  2. Module 2

    Fast feedback loops

    • Module-scoped builds and single-test runs in Maven and Gradle
    • Build cache, daemon and test selection for short cycles
    • Documenting the commands agents are allowed to run
  3. Module 3

    Architecture as a guardrail

    • Layer and package rules with ArchUnit
    • Spring conventions: configuration, profiles, dependency injection, transactions
    • Exercise: a feature across several layers, including review
  4. Module 4

    Tests worth having

    • Test slices with WebMvcTest and DataJpaTest
    • Integration tests with Testcontainers instead of in-memory shortcuts
    • The counter-check: would this test fail if the code were wrong?
  5. Module 5

    Upgrades

    • From javax to jakarta and other Spring Boot major-version changes
    • OpenRewrite recipes plus an agent for the remaining compile errors
    • JDK upgrades, deprecated APIs and dependency alignment

The build is the agent's teacher

An agent improves its code by compiling and running tests. If a full build takes many minutes, it waits, skips verification or runs into timeouts. The quickest win in most Java projects is therefore not a better prompt but documented commands: build one module, run one test class, start the application with a lightweight profile.

We work out these commands for your project and store them where every agent finds them, such as CLAUDE.md or AGENTS.md.

Upgrades: deterministic tools first, agents second

Large Java migrations work best as a combination. Deterministic refactoring tools such as OpenRewrite handle recurring patterns like namespace changes reliably and repeatably. Agents take on the irregular rest: custom code, failing tests and third-party libraries without a drop-in successor. Vendor upgrade features, for example in Amazon Q Developer, are compared neutrally.

  • Upgrade one dimension at a time: JDK, framework, libraries
  • Keep mechanical recipe changes and manual fixes in separate commits
  • Use the test suite and a post-deployment smoke test as the acceptance gate

For regulated environments

Many Java systems run in banks, insurers and public administration. We explain which code and context the tools send to their providers, how enterprise agreements differ from personal accounts, and how we run the workshop on a sandbox repository if your source code must not be shared with external tools. jocapps is ISO 27001 and ISO 9001 certified.

Frequently asked questions

Do we have to leave IntelliJ IDEA?

No. JetBrains AI Assistant, Junie and GitHub Copilot run inside the IDE, and terminal agents such as Claude Code or OpenAI Codex can work alongside it on the same project.

Does this work with older Java versions?

Yes. Agents are often most valuable in older codebases, for example to write missing tests or prepare an upgrade. We adapt the exercises to the Java and Spring versions you actually run.

What about Kotlin?

Kotlin on Spring Boot works with the same workflows. If your team writes Kotlin, we use it in the exercises and add Kotlin-specific conventions to the agent configuration.

Workshop

Get in touch

Which Java upgrade or test backlog should agents help you clear first?

Tell us roughly how many people will join and pick dates that work – we will get back to you within 24 hours with a proposal for content, format and schedule.

Prefer to talk first?

15 minutes, free and without obligation – with our AI Trainers and AI Code Engineers.

Book a call for free (opens in new tab) +49 761 519 751 50 (24/7)

Request a training / workshop

Workshop: Agentic Coding with Java & Spring Boot: Enterprise Workshop · 1 day

Dates that work for you (optional)

Suggestions about 2 weeks ahead – pick as many as you like

Reply usually within 24 hours · jocapps is ISO 27001 certified · Privacy