Chapiri Jaswanth Kumar

Pharma AI/LLM Insights & RAG Data Platform

– Architected an enterprise GenAI data ingestion pipeline processing 10M+ unstructured pharma sales call transcripts, clinical notes, and physician feedback logs into a high-dimensional vector database. – Orchestrated batch PySpark jobs to chunk texts, generate domain-specific embeddings via AWS Bedrock / HuggingFace, and populate a RAG (Retrieval-Augmented Generation) retrieval layer with sub-second vector search latency. – Enforced HIPAA-compliant PII redaction using AWS Comprehend Medical prior to embedding generation, reducing manual qualitative compliance review time by 80%.