Posted 1h ago
Loading this job…
Applied Machine Learning Intern
Posted 1h ago0 views
The problem you will work on
Every week MailerMen receives thousands of applications, and the ranking model that decides which five candidates a recruiter sees first is still a hand tuned scoring formula. You will replace it. Over this six month internship you sit with the matching team and build a learning to rank pipeline in Python, trained on two years of anonymised application and shortlist outcomes.
What the six months look like
You start by exploring the applications dataset in pandas and writing the feature engineering code for candidate skill overlap, seniority distance and location fit. From month two you train gradient boosted rankers, compare them against the current formula using NDCG and precision at five, and defend your evaluation choices in a weekly modelling review. By month four your model runs behind a feature flag on live traffic, and you own the dashboard tracking its recruiter click through rate.
Who this suits
You should be finishing a degree, or have just finished one, and be comfortable enough with Python that a notebook is somewhere you think rather than somewhere you copy from. No industry experience is expected. What matters is curiosity about why a model is wrong, and the stubbornness to keep asking until the data answers.
Responsibilities
- Explore two years of anonymised application and shortlist data in pandas and document what is actually usable
- Engineer ranking features covering skill overlap, seniority distance and location fit
- Train and tune gradient boosted learning to rank models in Python
- Benchmark every candidate model against the existing hand tuned formula using NDCG and precision at five
- Ship the winning model behind a feature flag and monitor recruiter click through rate
- Present modelling decisions in the weekly review and keep the experiment log current
Requirements
- Final year student or recent graduate in computer science, statistics, mathematics or a related field
- Working knowledge of Python and pandas from coursework, projects or self study
- Understanding of supervised learning basics including train test splits and overfitting
- Able to explain an evaluation metric and justify why you chose it
- Available for a six month internship with three days a week in Bengaluru
Skills
Benefits
- Monthly stipend paid on the first of every month
- One to one mentorship from a senior machine learning engineer
- Your model running on live recruiter traffic before you finish
- Internship certificate and a written performance reference
- Strong performers are considered for a full time offer