Building reporting services using Quarkus and AI
summary
Built a Quarkus middleware that exposes a simple users/orders API, uses Quarkus Dev Services + Postgres and import.sql for data, and plugs LLMs (local AMA/Llama3 plus OpenAI GPT‑4o) to translate natural‑language requests into SQL on the fly. The pipeline includes an LLM-to-SQL translator (prompted with schema, examples and retry logic) and an optional sensitive-field scrubber that calls a second model before returning results. It mostly works — some queries need reprompting, guardrails validate results, and the scrubber adds latency but successfully removes passwords.
AI-written summary · about
chapters (12)
- 0:00AI-powered reporting services
- 3:57Why Quarkus
- 7:08AI integration with LLMs
- 10:00Java models via AI prompts
- 15:16Sample data with AI
- 18:26Quarkus with PostgreSQL
- 20:13Dynamic SQL from natural language
- 25:17Testing AI queries
- 31:03Securing API with AI filtering
- 33:35Quarkus guardrails
- 35:25Future vision
- 39:05Q&A
related talks
36mCrafting Intelligent Cloud Native applications with Quarkus
50mDeveloping AI applications in Java with Quarkus and Langchain4j
01h 01mScaling AI with Java: Building Production-Ready LLM Applications
48mAgentic AI with Quarkus, LangChain4j and vLLM
47mQuarkus and AI Integrating Java and LLM to build intelligent applications
42mCrafting Intelligent Applications in Java with Quarkus / LangChain4j
47mCreating AI-infused Apps, the Easy Way 🤙
44mCrafting Intelligent Supersonic Subatomic applications with Quarkus