Teracnew this weekSWE Coding Benchmarks & RL Environments
Good fit for: anyone matching the study screening criteria.
Listing verified on the platform’s official board. Last checked September 25, 2026.
Remote · Contract · $45-75/hr, paid weekly · 10–20+ hrs/week
We build coding tasks used to train and evaluate frontier AI coding agents. We're looking for experienced software engineers to design them.
What the work involves
- Writing realistic engineering tasks on full-stack web apps: the kind of problem a real team has in its backlog, on codebases that grow as each task builds on the last.
- Writing the tests that decide whether a solution is correct.
- Going beyond the UI: new services and APIs, a backend in another language (Go, Rust, Java), gRPC, rate limiting — drawing on systems you've actually built.
- Sharing what you've seen work in other projects and flagging what should change in ours.
The apps are TypeScript (React, Node, SQLite, Playwright). Use AI tools as much as you like to write code and tests.
What we're looking for
- Solid professional software engineering experience, backend and frontend.
- You've built and debugged real systems, and you care about testing them well.
- Clear written communication, and comfort working independently from written docs.
- 10–20+ hours a week, starting soon.
How it works
A short screener, a brief AI interview, and a paid work trial on the codebase you'd keep working on. US and Canada preferred for timezone overlap; exceptional candidates elsewhere are welcome. LinkedIn and GitHub required.
Applying via Terac: Terac pays out as soon as your work on a study is approved. Expect screening questions per study — spots are limited and go quickly.