A SAI BHARATH

I specialize in AI agent systems and LLM backends, turning complex retrieval and reasoning into reliable, production-ready software. I’ve built Graph RAG pipelines and multi-agent workflows that surface accurate answers over real knowledge bases.

At NCR Atleos, I engineered LLM services on Azure OpenAI with prompt pipelines, retrieval validation, and failure-handling logic for production-grade internal Q&A workflows. I also containerised AI microservices with Docker and partnered with DevOps to improve monitoring and production reliability.

Before that, I supported backend services and internal tooling during my internship and helped standardize CI/CD with GitHub Actions, reducing manual deployment steps across dev and staging environments. My work consistently focuses on reliability, validation, and maintainable engineering practices.

On the project side, I built Dockwave to extract entities, construct a Neo4j knowledge graph, and enable context-aware querying across entire workspaces. I’ve also shipped an AI finance tracker with an LLM routing layer and deployed a real-time phishing detection system on AWS EC2 and a Chrome Web Store extension used by real users.