Principal Data Scientist

This is a remote position.

As the Principal Data Scientist at Migo, you will take full ownership of our credit risk modelling strategy, from the design and development of credit scoring, underwriting, and forecasting models to the end-to-end management of ML pipelines. You will research and implement cutting-edge modelling techniques, ensuring robust feature engineering, rigorous performance evaluation, and scalable deployment into production. Your work will directly drive lending decisions, financial performance forecasting, and market expansion strategies. Collaborating with cross-functional teams, you will generalise solutions for new markets, apply causal inference and statistical best practices, and maintain high standards in data integrity and model governance, helping Migo deliver responsible and data-driven financial products across Nigeria and beyond.

Responsibilities

• Full ownership of ML models and credit risk modeling methodology
  • Credit scoring models (probability of default, survival analysis models)
  • Underwriting models (credit line assignment and risk assessment models)
  • Forecasting models (projected financial contribution and performance metrics)
• Ownership of ML pipelines and workflows (ETL, data preprocessing, feature engineering, model training, model deployment)

Requirements

You are a good fit if you have:- A PhD in Computer Science, Statistics, Economics, Physics, or equivalent experience - Experience developing new modeling approaches, incorporating and adapting the latest methods in the field
- Experience developing ETL and feature engineering data pipeline
- Deep understanding of inductive biases of methods in ML and statistics
- Deep understanding of metrics and appropriate model performance measurement
- Ability to generalize product applications to new markets and partnerships
- Strong understanding of and experience with causal inference concepts such as positivity assumption, confounding, and exchangeability
- Deep understanding of ML and statistical concepts such as regularization, prediction vs. inference, multiple testing, cross-validation, boosting/bagging
- Experience with a variety of ML methods, especially those for tabular data and in the lending/finance space
- Experience with Python and object-oriented programming, developing and deploying production models, and contributing to shared, reusable libraries
- Experience with A/B testing
- Full proficiency in SQL

Originally posted on Himalayas