Snehal Garlapati

I evaluated and validated LLM outputs at Ethara.AI, focusing on accuracy, relevance, factual consistency, instruction adherence, and model reliability.

I improved model performance through prompt engineering, response analysis, data annotation, quality assurance, and actionable feedback for post-training and alignment workflows.

I built Luna, an AI desktop assistant with a FastAPI backend and React frontend, including a DocumentAI RAG pipeline for PDF and DOCX question answering. I also integrated LLM APIs, REST APIs, SQLite conversation history, and reusable AI services.

I use Python and SQL to turn data into useful decisions, from a sales analytics dashboard covering 1,000+ transactions to an NSE stock alert bot with automated Gmail alerts.