Data / ML Intern
Twelve models scaffolded. Two trained, both on synthetic data. So we don’t call it AI yet.
- Stipend
- ₹5,000 / month
- Duration
- 3–6 months (flexible)
- Location
- Remote — anywhere in India
Who can apply: Engineering, Statistics or Maths students and recent graduates
We've had more applications than we can read properly, and collecting more while people wait would be worse than saying so. The role isn't filled — check back, or leave your email on the careers page and we'll tell you the moment it reopens.
Where you’d start
Before any model comes the data. Six ingestion scaffolds exist and none has run end to end. You would take one — food, biomarker or drug — and turn it into a pipeline producing something worth training on.
What you’d do
- Clean and explore health datasets, and document what is in them.
- Help build training data pipelines for health prediction models.
- Train and evaluate baseline models with scikit-learn — and say so plainly when results are weak.
- Write up findings so a non-ML person can understand them.
What we need from you
This is the real bar. If you meet it, please apply — we would rather teach the rest.
- Python basics and some pandas or NumPy exposure.
- A first course in statistics or machine learning.
- Willingness to report a weak result. In health, an overstated model is worse than none.
Nice to have — genuinely optional
Not having these will not count against you.
- A Kaggle notebook, college ML project, or anything you have trained yourself.
- Some SQL.
- Interest in healthcare or bio data specifically.
What you’ll learn
Practical data cleaning, feature building, model evaluation, and the difference between a model that demos well and one that can be trusted near clinical decisions.
What we offer
- A ₹5,000/month stipend, paid monthly.
- Fully remote — work from anywhere in India, no relocation.
- Flexible 3–6 months, and we will work around your exam and semester dates.
- Real work on a live codebase or live content, not a sandbox project.
- A completion certificate and a letter of recommendation that describes what you shipped.
- Direct working contact with the founder — no layers, quick answers.
Being honest with you
We would rather you know this before applying than after joining.
- We are pre-launch and small. Some weeks are messy and priorities move.
- ₹5,000/month is a learning stipend, not a salary. Please weigh that honestly against your situation.
- We cannot promise a full-time offer at the end — that depends on funding. If it happens, you will hear it from us directly and early.
- You will need your own laptop and a workable internet connection.
How to apply
This is how it works when applications reopen.
- Fill in the form on this page — six questions and a file, about five minutes. No account needed.
- Attach your CV (a single PDF is perfect) and anything you have built — GitHub, Figma, a blog, a reel, a college project. Coursework counts.
- In three or four sentences, tell us why women’s and family health interests you. We read every one of these.
- No cover letter template needed. Plain, specific writing is better.
- Prefer email? careers@shely.health with the role in the subject line reaches the same place.
What happens next
- We read your application and reply either way, usually within a week.
- A 20–30 minute call to talk through your interests and answer your questions.
- One small take-home task, scoped to about 3–4 hours. We do not set unpaid multi-day assignments.
- A final call, then a decision.