Over two months, contributed to the kestra-io/kestra platform by developing extensible AI integration features and enhancing core platform capabilities. Built configurable LangChain4j plugin definitions for providers like Ollama, OpenAI, and Gemini, enabling runtime LLM service integration without code changes. Improved data discovery and filtering by launching a unified search and query system with advanced filtering for namespaces and labels. Established plugin scaffolding and CI/CD pipelines in TypeScript and Java, and refactored HTTP client usage across plugins for consistency. Addressed unit testing reliability and streamlined plugin development, resulting in a more flexible, maintainable, and AI-ready automation platform.