Posted 2d ago
MLOps Intern - Model Deployment and Monitoring
Posted 1mo agoBe an early applicant
What this role pays, in context
This listing does not state its pay. Data Scientist roles in India currently advertise a median of ₹24K a month.
25th pct ₹20.6K75th pct ₹27.1K
Median of 12 other live Data Scientist listings in India on this board that state pay. Roles that do not disclose a salary are excluded rather than counted as zero.
The gap we need closed
Models at MailerMen get trained well and deployed badly. There is no drift alerting, retraining happens when someone remembers, and nobody can tell you which model version scored a given candidate last Tuesday. Six months, one intern, that gap.
What you will actually do
You containerise the ranking model, put a small FastAPI wrapper in front of it, and get it serving behind our existing gateway. Then the monitoring: logging prediction distributions, comparing this week's input features against the training window, and firing an alert when the skill-tag vocabulary shifts because a new hiring trend arrived. The closing months cover the retraining schedule and a model registry simple enough that the next intern understands it in an afternoon.
Shape of the role
The position is unpaid and deliberately light on hours, two days a week from our Gurugram office with the rest flexible, plus a standing weekly session with the engineer who runs our infrastructure. It suits a final-year student or recent graduate who likes the plumbing as much as the modelling and has already deployed something small somewhere, even if only a hobby project on a free tier.
Responsibilities
- Containerise the ranking model and serve it behind a small FastAPI wrapper on the existing gateway
- Log prediction distributions and input feature statistics for every scoring request
- Build drift detection that compares live feature distributions against the training window
- Set up alerting so the team hears about a shifted skill-tag vocabulary before a user reports it
- Create a lightweight model registry recording which model version scored which request
- Define and document the retraining schedule, including what triggers an off-cycle run
Requirements
- Final-year student or recent graduate drawn to the infrastructure side of machine learning
- Python, plus enough Docker to build and run an image without following a tutorial line by line
- Has deployed at least one small service somewhere, a hobby project on a free tier counts
- Comfortable reading logs and dashboards to work out what broke and when
- Able to be in the Gurugram office two days a week
Skills
Benefits
- Certificate + Letter of Recommendation
- Deliberately light weekly hours so it fits alongside a final year of study
- Weekly mentoring from the engineer who runs our production infrastructure
- Hands-on access to the real deployment stack instead of a sandbox clone