Senior Machine Learning Engineer
Weekday AI · Mumbai, Maharashtra, India
Technology, Information and Internet · 11-50 employees
About the role
Design, build, and validate predictive and descriptive models using live and historical sports data. Collaborate with backend engineers to transition models from prototype to production while monitoring performance and accuracy.
What they look for
Requirements
Requires 5+ years of experience in building statistical or machine learning models in production environments. Candidates must possess strong Python and SQL skills, along with a deep understanding of applied statistics and data validation.
Full description
This role is for one of Weekday’s clients
Min Experience: 5+ years Location: Mumbai, Maharashtra, India JobType: full-time
This is a model-building role, not an ML platform or MLOps role. You'll be designing the statistical and machine learning models that sit at the core of our sports intelligence products — predictive models over live match data, and player and team performance metrics that are published, sold, and argued about by people who know the sport well.
Responsibilities:
- Design, build, and validate predictive and descriptive models over live and historical sports data - Define performance metrics and rating systems, and defend their methodology to product, commercial, and external stakeholders - Own feature engineering and the analytical datasets your models depend on, working from the team's data marts - Establish validation, backtesting, and monitoring practices appropriate to models whose outputs are published in real time - Take models from prototype to production, working with data and backend engineers on serving and integration - Investigate model failures and drift, and improve accuracy and robustness over time - Translate domain questions from product and sports stakeholders into tractable modelling problems
Requirements:
- 5+ years building statistical or machine learning models in production, with clear ownership of modelling decisions rather than only implementation - Strong grounding in applied statistics: probabilistic modelling, regression, time-series or sequential data, and rigorous validation - Strong Python and the standard modelling stack; strong SQL - Demonstrated ability to work with large, messy, real-world event data - Ability to explain and defend a model's methodology to a non-technical audience — a substantial part of this role is making a number credible, not just accurate - Judgement about model complexity: knowing when a simple, explainable model is the correct answer
Must-have skillsPython, SQL
