Strategic Project Lead – Software Engineering

This contract role puts an experienced operator in charge of the software-engineering data programs used to train and evaluate frontier AI models. It suits someone who has run coding data production before and can assess technical work as confidently as they manage people, budgets and deadlines. The engagement is with a company that supplies training data and research support to frontier AI labs.

What you'll do

• Run live coding data and evaluation programs, covering areas such as RLHF preference data, SWE benchmarks, agentic coding and code repair
• Recruit, calibrate and onboard contributors, and plan reviewer capacity
• Check code, tests and delivered artifacts, then decide whether work is accepted or sent back
• Monitor output, quality, rework, cost and deadlines, and build recovery plans when a program slips
• Report progress and risks to customer research teams, using Python or SQL for quality analysis and reporting

What you'll need

• Hands-on experience running coding data operations at production scale for frontier AI models
• Ownership of an entire production pipeline, from contributor calibration to cost control and getting late deliveries back on track
• Enough engineering depth to read code and tests and spot invalid work with little ramp-up
• A track record leading distributed engineers, reviewers and team leads

Good to know

• Three-month contract with an expected start the week after listing; no medical or paid leave included
• Full-time commitment of 8 hours a day, overlapping PST by at least 4 hours
• Fully remote
• Selection involves a roughly 20-minute AI interview and a 45-minute delivery interview