Geethanjali Baskar

I’m a results-driven Generative AI Engineer with 8.5+ years in IT and 3+ years building, fine-tuning, and deploying LLMs for real-world applications. I focus on turning model capabilities into reliable products through RAG, prompt engineering, and optimization strategies.

Most recently, I architected an enterprise production system using LangGraph to coordinate multi-agent teams end-to-end—automated code drafting, containerization, and continuous integration testing. I built self-correcting agent modules that execute tool calls in sandboxed environments, detect execution errors, and iteratively apply self-contained code patches.

I also developed high-concurrency Generative AI APIs with FastAPI, optimizing streaming tokens and structured JSON responses to support over 500,000 requests per day. On the research-to-product side, I built a hybrid RAG framework using Pinecone and Cohere reranking, reducing retrieval hallucinations by 38% for domain-specific documentation search.

Earlier, I strengthened scalable AI infrastructure on AWS—implementing cross-account IAM role handshakes across 6+ interconnected accounts to securely partition development and live workloads. I automated microservice stack deployment with Terraform, and integrated continuous deployment workflows with GitHub Actions and ArgoCD, while resolving production-critical Docker, Kubernetes, and PostgreSQL issues. I’ve also contributed as a software engineer and Appian developer, mentoring teammates and building workflow solutions aligned to best practices.