Machine Learning Foundations
Train, evaluate and explain a model on a public dataset, with a reproducible notebook.
Internship with[Partner company]- Duration
- 12 weeks
- Mode
- Remote
- Weekly effort
- 10–12 hours
- Level
- Intermediate
- Next cohort
- Feb 2027
- Fee
- ₹8,499
Overview
You will take a public dataset through the full modelling lifecycle — framing the problem, preparing the data, training and tuning a model, and explaining its results honestly, including where it's wrong.
The final deliverable is judged as much on the explanation and the reproducibility of your notebook as on the model's accuracy score — a model nobody can trust or rerun doesn't count as done.
Learning outcomes
- Frame a real-world question as a modelling problem
- Prepare and split data without leaking information
- Train and tune a baseline and an improved model
- Evaluate a model with the right metrics for the problem
- Explain model behaviour, including its failure cases
- Produce a reproducible, well-documented notebook
Project deliverables
Programme structure
- Orientation & problem framingMeet your mentor, choose the dataset, and frame the question as a modelling problem.
- Milestone 1 · Data preparationClean, split and document the dataset without leaking information across splits.
- Milestone 2 · Baseline modelTrain a simple baseline and evaluate it with the right metrics for the problem.
- Milestones 3–4 · TuningTune the model and compare it honestly against the baseline, including failure cases.
- Write-up & reproducibilityClean the notebook for reproducibility and draft the explanation of what the model does and doesn't do.
- Final project reviewSubmit the finished notebook and write-up. Completion decision within five working days.
About the partner company
[Two or three sentences about the company, its product or services, and the team interns will work alongside.]
Mentor guidance
Your mentor is a practitioner from the partner company. Their role is to review, question and guide — not to do the work for you. Our coordinators support you throughout.
Completion criteria
A certificate is issued only when every criterion below is met. You can track each one from your dashboard.
- Every milestone submitted and reviewed
- The minimum rubric score met on each milestone
- Final project review passed
- Original work, in line with the academic integrity policy
Certificate details
Issued digitally with a unique ID and QR code. It records:
- Your name, internship title and partner company
- Duration and completion date
- Skills assessed by your mentor
- A public verification link
Eligibility
- Who can apply
- Students currently enrolled in an undergraduate, postgraduate or diploma programme in India.
- Prerequisites
- Working knowledge of Python and basic statistics. Some exposure to pandas is expected.
- You will need
- A laptop that can run Python (or a free notebook environment), and a stable internet connection.