Posted 1h ago
Loading this job…
Machine Learning Intern (Python, scikit-learn)
Posted 1h ago0 views
Open to applicants anywhere
This is the one MailerMen internship that is not restricted to India. Work from any timezone as long as you overlap with Indian Standard Time afternoons twice a week. It runs for three months and is scoped tightly enough that a good intern walks away with something demonstrable rather than half a research project.
The problem
Job titles are chaos. "SDE 1", "Software Engineer I", "Sr. Softwr Engg" and forty thousand other spellings sit in our postings table with no canonical mapping. You will build the classifier that normalises them into a clean taxonomy of around three hundred roles. That means labelling a training sample by hand, comparing TF IDF against embeddings, measuring where the model fails, and shipping a batch job that tags new postings nightly.
How we support you
You get a weekly hour with our senior data scientist, a labelled starter set to build on, and honest feedback on your approach rather than a rubber stamp. Applicants straight out of university are welcome and no professional background is expected. Show us a notebook you are proud of and explain why you made the choices you made.
Responsibilities
- Hand label a training sample of job titles against a draft role taxonomy
- Build and compare TF IDF and embedding based classifiers for title normalisation
- Measure per class precision and recall and identify systematic failure modes
- Ship a nightly batch job that tags newly ingested postings
- Maintain a held out evaluation set and report accuracy changes over time
- Document the taxonomy and the model's known blind spots
Requirements
- Final year student or recent graduate in computer science, data science, mathematics or a related area
- Practical Python including pandas and scikit-learn
- Understanding of text vectorisation, classification metrics and cross validation
- Written English strong enough for asynchronous collaboration
- Able to overlap with Indian Standard Time afternoons at least twice a week
Skills
Benefits
- Monthly stipend paid in INR or the local equivalent
- Fully remote and open to applicants outside India
- Weekly hour with a senior data scientist
- Certificate of completion and a LinkedIn recommendation
- Freedom to publish a write up of your work
