James - Skilled Software Engineer for Ai Training Platform - Git Commit to Coding Task Conversion

We are seeking a skilled software engineer to convert real Git commits into high-quality, implementation-agnostic coding tasks for an AI training platform. The role involves analyzing code changes, writing clear technical specifications, and creating end-to-end tests that validate the described behavior. Ideal for someone who enjoys digging into complex codebases, reading diffs, and producing precise, testable specs. Responsibilities: Transform selected commits into complete coding tasks. Write unambiguous, natural, implementation-agnostic task descriptions. Capture every logical requirement the tests must verify. Extract correct code changes and remove unrelated modifications. Ensure the golden patch compiles and passes all new tests. Write end-to-end, implementation-agnostic tests covering all described behavior. Ensure the base commit fails at least one test and the golden patch passes all. Follow required naming conventions and iterate through QA as needed. Deliverables: Task name, merge hash, base hash. Fully specified task description. Optional base patch. Clean golden patch. Test patch. Explanation of expected vs incorrect AI behavior. Workload: Tasks require 2–10 hours depending on complexity. Ongoing work available for strong performers.