You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm a Senior Data Scientist at Genpact, building production GenAI, agentic AI, and NLP systems — Text-to-SQL platforms, LLM extraction pipelines, and semantic search that ship real, measured impact instead of notebook demos.
🚀 What I'm building at Genpact & UST
Conversational Analytics (Text-to-SQL) Platform — agentic LangChain/LangGraph system over Azure OpenAI, cutting query-formulation time by 65% for 50+ beta users across 10 concurrent sessions, scaling to 500+ weekly queries at a 92% task-completion rate with 12+ cross-functional teams.
Invoice Processing Pipeline — LangChain + Azure OpenAI extraction with human-in-the-loop review, reaching 85%+ accuracy across 10,000+ invoices in 8+ languages and cutting manual processing time by 60%.
Provider Search System (UST) — NER + semantic/vector search mapping layperson language to medical terms, boosting search efficiency by 30% and accelerating data prep by 40% — my most-cited career win.
Healthcare Analytics (UST) — BigQuery-based patient-adherence analysis and an anomaly-detection system for categorical healthcare data, built HIPAA & GDPR compliant.
LLM-Lora-PEFT_accumulate explores optimizations for Large Language Models (LLMs) using PEFT, LORA, and QLORA. Contribute experiments and implementations to enhance LLM efficiency. Join discussions …