Senior Software Engineer: C++ (LLM Evaluation & Repository Validation)

What you'll do

Turing is building evaluation and training datasets from the public history of real repositories, with a human checking each step. In the C++ strand of that project you pick well-maintained GitHub projects, work out which issues are genuinely hard for a language model, and make the repository runnable so a fix can be verified.

Day to day that means:

• Triaging issues across trending open-source C++ libraries
• Setting up the repository, including Dockerisation and environment configuration
• Judging unit test coverage and quality, since a weak test suite cannot tell a good fix from a bad one
• Modifying and running the code locally to see how an LLM performs on bug-fixing scenarios
• Working with researchers on which repositories and issues are worth including

The ad adds that there are opportunities to lead a team of junior engineers.

Who fits

• At least 3 years of overall experience, with strong C++ (the ad asks for strength in at least one language, and for this listing that language is C++)
• Comfortable with Git, Docker and basic pipeline setup
• Able to navigate large, unfamiliar codebases and run, modify and test real projects locally
• Open-source contribution or evaluation experience is a plus
• Nice to have: earlier LLM research or evaluation work, or experience building developer tools or automation agents

The ad frames the role as tech-lead level, so treat the 3 years as a floor, not a typical profile.

What it pays

The ad does not publish a rate. Turing sets rates per project; see Turing for how that generally works.

Commitment and location

• Three time options: 20, 30 or 40 hours a week, with at least 4 hours a day and 4 hours of overlap with Pacific time
• Contractor assignment, with no medical or paid leave
• The ad lists India, Pakistan, Nigeria, Kenya, Egypt, Ghana, Bangladesh, Turkey and Mexico as the eligible locations
• The ad states no contract length for this C++ listing

Worth knowing

Good:

• Real engineering work (Docker, build systems, test suites) rather than abstract labelling
• C++ repositories are a narrower niche than Python ones, so the pool of qualified applicants is smaller
• Fully remote, with a choice of weekly hours
• A route to team-lead work is mentioned

Less good:

• No pay figure is published
• Four hours of PST overlap is demanding from South Asia or East Africa
• Eligibility is limited to the countries named
• Contractor terms mean no benefits or paid leave

How hiring works

The ad describes about 75 minutes of evaluation across two interviews: a 60 minute technical round and a 30 minute technical and cultural discussion.

About this listing

Read from Turing's job board on 18 October 2026. The hours, locations, interview format and employment type are Turing's own text. See Turing.