BHAVANA KRISHNAN

About

LLM engineer who has deployed a production RAG system (LangChain/FAISS/FastAPI) for enterprise semantic search, tackling real-world latency by LoRA/PEFT fine-tuning and grounding outputs with retrieval. Brings strong MLOps (Docker, AWS EKS, CI/CD, MLflow) plus stakeholder-facing explainability experience using SHAP to align ML-driven financial guidance with non-technical domain experts.

Experience

Education

Key Strengths

• Built and deployed a scalable RAG semantic search system for enterprise documents (LangChain/FAISS/FastAPI)
• Reduced inference latency via LoRA/PEFT fine-tuning with adapter merging for zero inference overhead
• Grounded LLM responses with FAISS-based retrieval over transactional data to improve consistency
• Strong MLOps/orchestration across Docker + AWS EKS with CI/CD (GitHub Actions, Jenkins) and MLflow model/version tracking
• Designs reliable agent/workflow evaluations with tracing/observability, regression tests, and CI-integrated metrics
• Effective cross-functional collaboration translating finance rules into ML features and using SHAP to explain model behavior