Headline

Data Scientist & GenAI Engineer | LLM Fine-Tuning · Agentic AI · RAG Pipelines | Snowflake Cortex · LangGraph · Vertex AI | IIT Madras M.Tech | Pharma & Healthcare AI

About

I build production-grade AI systems that actually ship — from LLM classification engines processing thousands of records at Simtra BioPharma, to a multi-agent ecosystem with 10+ specialized agents at Fluence. Currently a Data Scientist at KPI Partners, I architect end-to-end GenAI and Agentic AI platforms for enterprise clients across pharma, energy, and healthcare verticals. 🔧 What I work with: • LLM Fine-Tuning, RAG Pipelines, Multi-Agent Orchestration • Snowflake Cortex, LangGraph, LangGraph, Vertex AI, Gemini 2.5 • PyTorch, HuggingFace, LangChain, Google ADK • GCP, Azure, FastAPI, Streamlit, GitHub Actions CI/CD 📌 Recent highlights: • Built a pharmaceutical quality intelligence platform predicting GMP deviation categories with confidence scoring and human-in-the-loop review • Architected a Pre-Handover Engineering Agent enabling 7-parallel-task document retrieval with self-improving SME approval loop (grounding ≥80%) • Developed Text-to-SQL platform hitting 88% accuracy across 500 benchmark questions 🎓 M.Tech, Clinical Engineering — IIT Madras (CGPA: 8.83) 📄 Published in Springer Q2 Journal (2026) | Microsoft Certified Azure Data Scientist (DP-100) I'm always curious about applied AI problems at the intersection of engineering and impact. Let's connect!

Location

Chandrakona

AI stack

antigravity

cursor

gemini

hugging face

langchain

claudecode

Work experience

Data Scientist at KPI Partners, Jul 2025 - Present · 1 yr 2 mos, Bengaluru, Karnataka, India, Architecting and deploying production-grade AI and GenAI platforms for Fortune-500 enterprise clients. • Design and ship LLM-powered systems for intelligent automation, knowledge retrieval, and decision support across pharmaceutical, energy, and enterprise verticals using Snowflake Cortex, LangGraph, and Vertex AI. • Lead AI governance implementation: auditability controls, confidence scoring, and human-in-the-loop review across multi-agent systems with 46+ specialized agents. • Establish prompt engineering versioning standards and CI/CD pipelines (GCP + GitHub Actions) ensuring high availability and reproducibility across agentic ecosystems. • Build production batch inference pipelines processing thousands of deviation records with approval workflows and audit trails.

ABM Trainee at HighRadius, Jul 2022 - Mar 2023 · 9 mos, Bhubaneshwar, Odisha, India

Education

M. Tech, Clinical Engineering, Indian Institute of Technology, Madras

M.Tech, Clinical Engineering, Sree Chitra Tirunal Institute for Medical Sciences and Technology

Public links

github: https://github.com/tufan-2000