We run a marketplace for climate projects. Buyers want to understand delivery risk before they commit, but our current risk score is basically a spreadsheet maintained by two people.
We need a data scientist to design a first proper project risk scoring model using project history, verification delays, geography, partner track record, and delivery milestones.
This needs to be transparent enough for buyers and internal ops teams. We are not looking for a black box.
Requirements
Technical stack needed for this mission.
- Experience building interpretable scoring models
- Strong data cleaning and validation skills
- Comfortable with sparse, messy operational data
- Can deliver documentation and a handover walkthrough