Sanjay Neela

Architected and delivered enterprise healthcare RAG systems covering ingestion, chunking, metadata enrichment, embedding generation, vector indexing, hybrid search, and grounded response generation. Built secure asynchronous FastAPI services and integrated Azure OpenAI for clinical summarization, question answering, and workflow automation. Developed hallucination mitigation, response validation, token optimization, LLM evaluation, monitoring, RBAC, and HIPAA-aligned security controls. Containerized workloads with Docker and deployed scalable services to Azure Kubernetes Service with autoscaling and background embedding workers.

Designed Python and FastAPI backend systems for ML inference and NLP workloads, supporting high-volume financial applications. Implemented semantic search using embedding-based similarity, created preprocessing pipelines for feature extraction and embedding preparation, and built REST APIs for downstream banking systems. Improved inference latency and throughput through Redis caching and asynchronous execution. Standardized deployment of containerized ML services using Docker and Kubernetes.

Built end-to-end machine learning pipelines for data ingestion, preprocessing, feature engineering, model training, and AWS deployment. Developed NLP-based product classification and embedding-based search ranking improvements to enhance retail catalog discoverability. Used vector indexing and Redis caching to improve search performance and implemented CI/CD pipelines to automate model deployments and accelerate production releases.