Headline

Applied AI Engineer @ IF MedTech | Agentic AI, RAG, LLM Systems & Evals | LangGraph, LangChain, PyTorch | CV pipelines serving 270k+ users | Python, FastAPI, Next.js

About

I build AI systems that survive production. I work across the LLM and computer vision stack. I built MedRAG, an agentic RAG assistant over 220+ clinical documents (LangGraph, Pinecone), and the eval harness around it (RAGAS, LangSmith) that took retrieval recall@5 from 0.70 to 0.87 and faithfulness from 0.72 to 0.89. I also own a multi-model CV inference pipeline (Faster R-CNN, U-Net) that has served 270,000+ unique users, and I've fine-tuned and INT8-quantized a ViT-Base/16 to run on-device at ~0.7s per image. What I care about most is making AI measurable: eval harnesses, golden datasets, regression runs in CI. Plenty of engineers can wire up an agent; fewer can tell you whether it actually got better. Around the AI core I ship full-stack: FastAPI and Node.js backends, Next.js frontends, Flutter and React Native apps, deployed on AWS with Docker and CI/CD. I've also built and shipped products solo end to end as a freelancer, from first commit to production. Open to AI Engineer and Applied AI roles: remote (India or global) or on-site in Bengaluru, Hyderabad, or Gurugram. mayankbaranwal95@gmail.com

Location

Bengaluru

AI stack

claude code

codex

hermes agent

langchain

langgraph

mcp

Work experience

Applied AI Engineer at IF MedTech Pvt Ltd, Jun 2025 - Present · 1 yr 4 mos, On-site

Software Development Engineer at Ripple Healthcare, Jan 2024 - May 2025 · 1 yr 5 mos

Software Development Engineer at FlipTree Technologies, Jul 2023 - Dec 2023 · 6 mos, Remote

Education

Foundations of Modern Machine Learning, International Institute of Information Technology Hyderabad (IIITH)

Bachelor of Technology - BTech, Bundelkhand Institute of Engineering and Technology, Jhansi

Public links

github: https://github.com/mayankbaranwal0

mayankbaranwal.vercel.app: https://mayankbaranwal.vercel.app/