AI / Machine Learning Engineer
Build and ship the models behind NADI ai and KisanMitra, end to end, to people who cannot afford a paid alternative.
- Ref
- J01
- Team
- Platform
- Where
- Remote, anywhere in India
All roles
HCF runs free software for people the commercial market ignores: diagnostic support where the nearest radiologist is a day away, and crop advice for farmers who do not read English. Both are held up by machine learning that has to work on bad connections, small data and real stakes.
You would own pieces of that end to end — data in, model out, deployed, and then watched to see whether it actually helped.
What you would do
- Build and evaluate models against real, messy data rather than clean benchmarks
- Take a model from notebook to a deployed service that survives production
- Instrument what you ship, so failure is visible instead of assumed
- Write down what did not work, so the next person does not repeat it
What we are looking for
- Python, and comfort with PyTorch or TensorFlow
- You have trained something on data you had to clean yourself
- Enough backend understanding to serve a model over an API
- Willingness to say "this model is not good enough to ship yet"
These are not calculator projects or website clones. You will work on RAG systems, real databases and production AI pipelines — the kind of work you can actually talk about in an interview.