Bachelor Project
May 2024A sophisticated multi-agent system designed to automatically generate functional REST and GraphQL API interfaces using Large Language Models.
Key Features
🤖
Multi-Agent Architecture
Built using Google Gemini and an LLM Provider Factory pattern. Deploys specialized agents to analyze requirements and generate code for REST/GraphQL APIs.
âš¡
Real-time Streaming Backend
Architected the backend with FastAPI and MongoDB, utilizing Server-Sent Events (SSE) to stream generation progress and agent activities to the frontend in real-time.
🚀
Automated DevOps
Integrated automated deployments via Genezio for serverless scaling. Implemented Langfuse for detailed observability, tracing, and cost analysis of LLM usage.
Technologies
PythonFastAPIReactMongoDBGeminiAWSDockerGenezioLangfuse