AI Intern at Stealth startup

Built a Retrieval-Augmented Generation (RAG) system with Azure OpenAI and Ollama, integrating HyDE, step-back prompting, and query analysis for 10–15% performance gain. Developed Hybrid Search and persistent Chat Memory using Qdrant, Weaviate, and Mem0 with metadata-enriched retrieval pipelines. Designed Router Agent for multi-task LLM routing via LiteLLM and containerized deployment with Docker for production readiness.