AI-Driven Development Meets Jakarta EE

As AI tools increasingly assist in software development – from code generation to architecture decisions – the value of stable, well-documented, and consistent APIs has never been greater. In this talk, we’ll explore how Jakarta EE and MicroProfile provide the perfect foundation for AI-assisted and even autonomous application development. Drawing on practical experiments developing applications with the help of AI, I’ll demonstrate how modern AI-enhanced development tools can understand and generate Jakarta EE applications more effectively than ad-hoc or framework-specific codebases. Producing up-to-date code that works together much better than if assembled from various libraries or frameworks. We’ll also discuss the role of open standards in making application behavior predictable for AI, improving maintainability, testability, and avoiding vendor lock-in. You’ll leave with a clear picture of how enterprise Java can thrive in an AI-assisted future – and how your team can start leveraging AI tools today to boost productivity while staying grounded in proven architectures.

More info will be added soon (code examples, example prompts, ...)

Transcript

1. Ondro Mihalyi @OndroMih AI-Driven Development Meets Jakarta EE Why Standards Matter More than Ever
2. @OndroMih https://omnifish.ee The AI Development Challenge Why AI Struggles with Inconsistent Patterns AI excels with clear, documented patterns Fragmented ecosystems create confusion and errors Inconsistent structures reduce code quality Standards provide the clarity AI needs
3. @OndroMih https://omnifish.ee Framework Fragmentation Problem Why Scattered Libraries Confuse AI Tools Multiple ways to solve the same problem Inconsistent documentation across libraries AI generates conflicting or incompatible code Maintenance nightmare with mixed approaches
4. @OndroMih https://omnifish.ee Ad-hoc vs. Guided AI Code public class Fibonacci { public static void main(String[] args) { int n = 10; // number of terms int a = 0, b = 1; System.out.print(a + " " + b + " "); for (int i = 2; i < n; i++) { int next = a + b; System.out.print(next + " "); a = b; b = next; } } } // Generates fibonacci sequence up to n terms using iterative approach // for memory efficiency and O(n) time complexity public static void generateFibonacci(int n) { if (n <= 0) return; long firstNumber = 0, secondNumber = 1; if (n >= 1) System.out.println(firstNumber); if (n >= 2) System.out.println(secondNumber); for (int i = 3; i <= n; i++) { long nextNumber = firstNumber + secondNumber; System.out.println(nextNumber); firstNumber = secondNumber; secondNumber = nextNumber; } }
5. @OndroMih https://omnifish.ee Standards-Based Architecture Predictable Patterns AI Can Understand Eclipse specifications written in natural language Clear intent behind every API design decision Consistent patterns across all Jakarta EE APIs AI can reference comprehensive specification documents
6. @OndroMih https://omnifish.ee Annotation Semantics with Context Specifications Explain Purpose and Behavior @Inject, @Path, @Transactional have well-defined meanings Specifications describe when and how to use annotations AI understands context, not just syntax Self-documenting code through meaningful annotations @ApplicationScoped public class SchoolService { @PersistenceContext(unitName= "peoplePU") EntityManager entityManager; @Inject ValidationService validation; @Transactional void initializeSchoolData() {
7. @OndroMih https://omnifish.ee Comprehensive Use Case Coverage Specs Include Real-World Scenarios Business scenarios documented in specifications Common patterns and anti-patterns explained AI learns from proven enterprise use cases Reduces trial-and-error development
8. @OndroMih https://omnifish.ee Mature Ecosystem Integration CDI, Jakarta REST, MicroProfile Working Together APIs designed to complement each other Dependency injection connects everything seamlessly AI understands how components interact No integration guesswork required
9. @OndroMih https://omnifish.ee Specification Stability Long-Term Predictable APIs Backward compatibility guarantees Evolutionary, not revolutionary changes AI training data remains relevant over time Investment in learning pays long-term dividends
10. @OndroMih https://omnifish.ee Platform and Profile Specifications APIs Work Together, Covered by tests Platform and Profiles document integration with TCK tests Cross-specification integration patterns documented AI knows how Jakarta REST works with CDI MicroProfile builds on Jakarta EE foundations
11. @OndroMih https://omnifish.ee Implementation Alignment and TCK Code Examples That Match Specification Intent Technology Compatibility Kit ensures consistency Compatible implementations follow specifications exactly AI learns from verified, compliant code examples Reduces implementation-specific quirks
12. @OndroMih https://omnifish.ee Security by Design Jakarta Security Security specifications integrated with platform AI applies security patterns consistently (when asked) Authentication and authorization built-in Reduces security vulnerabilities User/password OpenID Connect Certificates JWT Bearer Auth Custom MicroProfile JWT
13. @OndroMih https://omnifish.ee Testing-Friendly Architecture CDI and Dependency Injection Advantages Dependency injection enables easy mocking AI generates testable code by default Specifications include testing guidance Unit and integration testing patterns documented
14. @OndroMih https://omnifish.ee AI-Friendly Code Patterns Reference specifications in AI prompts Use standard project structures (Maven archetypes) Leverage specification examples and patterns Ask AI to generate documentation and visual diagrams
15. @OndroMih https://omnifish.ee Team Productivity Tips Train team on specification fundamentals Create AI prompt templates referencing specs Build prompt templates to review best practices Develop specification-compliant project templates with AI instructions
16. @OndroMih https://omnifish.ee Thank You! Questions? Jakarta EE Consultancy (migrations, tuning) Enterprise Application Development GlassFish Server Support Jakarta EE Training OmniFish - Jakarta EE Consulting & Support https://omnifish.ee