turning ideas into deployed ml systems
I build ML systems that ship — from CNN-based classifiers to RAG pipelines that cite their sources. Model development, experiment design, and getting it into production with Docker and FastAPI, end to end.
I'm an AI/ML engineer who turns ideas into working systems — from CNN-based image classifiers to RAG pipelines that actually cite their sources. My experience spans the full lifecycle: model development with CNNs and Transformers, careful experiment design, and shipping models into production with Docker and FastAPI.
I'm comfortable owning a project end-to-end — from messy data and a blank notebook to a containerized service running in the real world. What drives me is practical problem-solving, where solid engineering is treated as a first-class part of the model, not an afterthought.
See it in action →A fully local RAG application to chat with PDF documents — embeddings and generation run entirely on-device via Ollama, with no cloud or API dependencies.
A production-grade visual search engine for e-commerce that finds visually similar products across a catalogue using CLIP embeddings and vector search.
An NLP-based Q&A system that retrieves answers from custom datasets and Wikipedia using transformer embeddings for semantic search.
A real-time multiplayer backend for the card game Call Break, built on asynchronous WebSockets with Redis-backed state for smooth concurrent play.
B.Tech. in Computer Science Engineering (Honours)
Punjab, India
Whether it's building ML pipelines, deploying models to production, or solving a gnarly data problem — I'd love to hear about it. I typically respond within 24 hours.
ashishbhandari365@gmail.com