About the role
This is a hands-on, build-first role. You will own ML problems end-to-end: defining the problem, experimenting, building the model and taking it to production. Your models will directly drive how users convert, which lender they are matched with, and how repayments are managed. There is no legacy system to maintain. Most of what you build will be 01
Responsibilities:
- Build and optimise predictive models for customer behaviour, conversion, credit risk, collections and repayment.
- Design features and modelling frameworks on real lending and credit-bureau data.
- Run feature engineering, model validation, experiments and performance tuning.
- Build GenAI/LLM applications where they solve a real problem (RAG, embeddings, NLP).
- Work with Product and Engineering to take models from prototype to production.
- Spot and pitch new AI/ML use cases across the business.
Requirements:
- 2-3 years of hands-on experience in Data Science or Machine Learning, building models on real business data (not only academic projects).
- Strong Python and advanced SQL.
- Solid grounding in ML, statistics, predictive modelling and feature engineering.
- Hands-on with Scikit-learn, XGBoost, LightGBM, PyTorch or TensorFlow.
- At least one model you've taken to production or into a live business decision.
- An ownership mindset: you've built things without a detailed spec.
Good to Have:
- Experience in NBFC, lending, BFSI or fintech (credit risk, collections, underwriting, propensity or fraud models).
- Exposure to GenAI, LLMs, NLP, embeddings or RAG.
- Startup experience.