01
LLM-powered Document Search Engine
August 2025
- Developed an LLM-powered document Q&A system using Python, FastAPI, and a microservices architecture.
- Utilized Weaviate vector database coupled with OpenAI vectorizer for efficient semantic search, enabling the system to retrieve contextually relevant document chunks.
- Architected a decoupled backend using RabbitMQ for asynchronous document processing, ensuring high availability.
- Utilized Docker Compose for containerizing services (Weaviate), simplifying local development and deployment.