SUJAY Kanakamedala

About

Built and deployed an enterprise RAG application at Centene to help clinical teams retrieve insights from large internal policy document sets, cutting manual research by 30–40%. Implemented custom domain-adapted embeddings (SageMaker + BERT transfer learning) and hybrid retrieval (BM25 + Pinecone) to drive a 22% relevance lift, and ran the system in production on AWS EKS with CI/CD, MLflow, and Prometheus monitoring (99% uptime, ~40% latency reduction).

Experience

Education

Key Strengths

• Architected and deployed enterprise RAG system for clinical policy retrieval
• Reduced manual research effort by 30–40% across business units
• Improved retrieval relevance by 22% using custom embeddings and hybrid retrieval (validated via BLEU and human review)
• Mitigated hallucinations and improved domain-specific accuracy for clinical jargon
• Production-grade deployment on AWS EKS with autoscaling, rolling updates, and 99% uptime
• Achieved ~40% latency reduction through containerized FastAPI services and orchestration
• Strong MLOps discipline: MLflow experiment tracking/model versioning, CI/CD, retraining pipelines
• Comprehensive testing strategy: unit/integration tests (Pytest) and load testing (Locust) with monitoring (Prometheus)
• Effective model/retrieval/prompt selection via benchmarking (GPT vs fine-tuned Llama 3), RAGAS scoring, and A/B tests
• Proven stakeholder collaboration with non-technical clinical policy teams via workshops and iterative demos (Streamlit)