Open to opportunities

Madan khatri

@madankhatri

Message

I build low-latency AI applications, MLOps pipelines, and full-stack ML products.

Nepal

Message

What I'm looking for

I'm looking for remote internships, freelance contracts, and collaborative AI research opportunities where I can build production-ready AI applications, reproducible ML pipelines, and scalable full-stack systems.

I've built production-ready AI applications spanning enterprise RAG, real-time sentiment analysis, visual speech recognition, and cardiovascular risk prediction.

For an Enterprise RAG Chatbot System, I architected an asynchronous FastAPI backend with Pinecone semantic search, MySQL document metadata, Redis chat memory, and Groq Cloud for low-latency LLM inference. I also developed a Chrome Extension and containerized Flask API for real-time YouTube comment analysis, supported by a reproducible five-stage DVC and MLflow pipeline.

My deep learning work includes a TensorFlow/Keras lip-reading system using Conv3D, bidirectional LSTMs, and CTC loss to decode speech from video frames.

With CardioSense AI, I trained a cardiovascular risk model on 70,000 patient records, reaching an AUC of 0.8003 and improving 9.6% over baseline. I maintain 35+ active open-source repositories and am open to remote internships, freelance contracts, and collaborative AI research opportunities.

Experience

Work history, roles, and key accomplishments

AI/ML Engineer & Full-Stack Developer

Self Employed

Highly driven AI/ML Developer with 4 years of hands-on experience designing, building, and deploying production-ready, low-latency end-to-end AI applications. Proven expertise in combining deep learning, computer vision, and generative AI with robust backend architectures and MLOps practices.