Lokesh Chowdary Unnam

I'm building LLM-powered applications, backend services, and multi-agent AI systems, with a focus on scalable AI infrastructure and reliable deployment.

At Evolutyz Corp, I redesigned credit deduction to occur after agent execution, improving billing accuracy across distributed failure scenarios. I also built a backend reference implementation for a Performance Management System, including scalable APIs, business logic, and database workflows.

I'm integrating Langfuse observability to capture LLM traces, latency, token usage, costs, and execution logs for production debugging.

My projects include a multi-modal RAG system with validated ingestion, vector search, and retrieval safeguards, plus Khetsaathi, a multilingual voice AI assistant that helps farmers diagnose crop diseases and receive location-specific seven-day treatment plans. I earned first place at the AI for Vizag Buildathon for an AI-driven Answer Engine Optimization solution.