Turing

Turing is a San Francisco-based technology company founded in 2018 by Jonathan Siddharth and Vijay Krishnan. It started as a way to connect software developers with remote roles and now also supplies specialist talent and technical work for advanced AI development. Its services cover software engineering, AI model improvement and research support. You can read the founders’ account of the company’s early direction in Turing’s official founding story.

Turing operates a global talent network rather than a single type of job program. That distinction matters: a long-term engineering placement, a short expert project and an AI research contract can have different selection steps, hours and contract terms.

AI data training work at Turing

Turing recruits software engineers, researchers, PhD-level specialists and professionals with strong knowledge of subjects such as mathematics, physics, chemistry, biology and other technical fields. AI projects can involve writing or reviewing difficult prompts, correcting model answers, checking code, building benchmarks and creating reinforcement-learning environments. A benchmark is a structured test used to measure an AI system; a reinforcement-learning environment is a controlled setting in which a model practices a task and receives feedback.

The company’s current Turing Advance overview describes work on post-training data, coding and reasoning, multimodal tasks and AI safety. “Post-training” means improving a model after its initial broad training, often with carefully written examples, expert evaluation and feedback. Other Turing opportunities are conventional remote software or machine-learning roles, so read the individual listing rather than assuming every opening is data-labeling work.

What is it like to work at Turing?

Many AI specialist opportunities are remote contracts, although the legal arrangement and eligible locations depend on the role. You may work as an independent professional on a defined project, while other parts of Turing advertise full-time or longer-term placements. Hours can range from flexible part-time commitments to a normal working week, and some projects require overlap with a client’s time zone.

Being accepted into Turing’s network is not a guarantee of paid work. Your profile and assessment results are used to match you with current demand, and there may be gaps while no suitable project is available. Matching can also depend on technical seniority, subject expertise, language, country and availability. For roles that let you state an expected rate, the final amount still needs to be agreed for the engagement.

Application process at Turing

You normally start by creating a profile and uploading a résumé. The next steps depend on the work. Developer candidates may complete a seniority questionnaire, technical multiple-choice tests, a coding challenge and a live technical interview. Domain-expert and AI-training roles can instead use subject tests, writing or reasoning assessments, and an interview focused on your specialist experience.

Identity, location and work-history checks may be required before onboarding. If you pass the general evaluation, Turing can make your profile available for matching; it does not mean a project starts immediately. A client or internal team may conduct another interview, and you will receive project-specific terms before accepting. 

Sources: Company website

Information last updated August 2026.