Lead Data Scientist
About the Role
Own the data and risk function behind Credible, an early-stage working-capital product built on the Payable network. You'll decide how every advance is scored and sized, using card-payment data across the network, and take it from rules-based decisioning to a trained model in production.
This is an ownership role, not a people-management position. You'll take responsibility for turning data into decisions, from outcome labels and backtested rules to trained models used in production. The engineering lead owns the platform and data ingestion. Your responsibility starts from there.
WHAT YOU'LL OWN
- The data register and outcome labels that make scoring and evaluation reliable.
- Scoring and renewal rules, including versioning, backtesting and monitoring.
- Early-warning signals that operations can use to act on risk.
- The move from rules-based decisions to a trained model in production.
WHAT WE'RE LOOKING FOR
- At least one machine-learning model you've owned end to end in production, with real decisions depending on it.
- The ability to write production Python and SQL yourself.
- Daily, hands-on use of AI coding tools. If your employer restricts them, be ready to show how you use them in your own work.
- Accountability for outcomes, sound judgment and the ability to work in an early-stage business.
- Around five or more years of relevant experience is a guide, not a fixed cutoff.
- Experience in risk or decision science within fintech, payments or technology is a strong fit.
- Bank credit-risk experience can fit if you're still a hands-on coder and use AI tools.
This is not a research-only, reporting or BI role. We want someone who has been responsible for a live model after deployment.
WHAT SUCCESS LOOKS LIKE
- Within six months, you'll own scoring and renewal decisions, have versioned rules with backtests and a live data register, and early-warning signals in operational use.
- Within twelve months, a first trained model in production.
PRACTICAL DETAILS
- Based in Colombo 3, four days in office, one flexible.
- Laptop and paid AI tools provided.
- Target start February-April 2027.
- Sri Lankans overseas ready to return are welcome.
Technologies
- Python
- SQL