We are hiring a Quant / Data Scientist to build the model layer behind Mettle Data. This role is focused on developing and deploying advanced predictive systems across sports and market data, with a particular emphasis on deep learning approaches alongside traditional statistical models.
You will work on end-to-end model development: feature generation from noisy, high-frequency data, training and tuning deep learning architectures, backtesting, calibration, live-versus-backtest diagnosis, and production monitoring. A core part of the role is improving the live evidence loop and determining which modeling approaches (including neural network-based pipelines) should be advanced into production.
We already operate a profitable model in production; the next phase is pushing model sophistication while maintaining rigorous validation and real-world performance discipline.
We are looking for strong Python and SQL skills, deep understanding of machine learning and modern deep learning techniques, and a high bar for empirical validation. Experience working with real-world, messy datasets is essential. Backgrounds in sports, quantitative finance, or trading systems are a strong plus.
This is an opportunity to join a small, highly-skilled and ambitious team, with significant opportunity for progression. This role is fully-remote.