Software Engineer for AI Training and Evaluation

Role Overview

Help shape how frontier AI systems are trained and evaluated by applying production software engineering judgment to real-world code, designs, requirements, and failure modes. This is an ongoing talent pool for AI training and evaluation projects, rather than a single immediate opening. Projects may begin on short notice, and candidates are contacted when a project matches their technical background.

Key Responsibilities

• Create engineering problems based on systems you have shipped, including requirements, constraints, failure modes, and reference answers used to assess model outputs.
• Evaluate AI-generated code and technical designs for edge-case handling, appropriate complexity, and whether tests validate the behavior they claim to test.
• Identify code that may pass review but fail in production because of concurrency, scale, error handling, or operating costs.

Qualifications

• Professional experience building and shipping software.
• Strong written communication and the ability to clearly explain technical reasoning.
• Comfort working through ambiguous tasks and flagging unclear instructions.
• Experience owning production code and addressing its real-world behavior is valued more than training background.

Technical Coverage

• Frontend engineering, including React, TypeScript, and Next.js.
• Backend engineering, including Java, Go, and Node.js, as well as full-stack development.
• Mobile development for Android and iOS, embedded systems, and IoT.
• DevOps, platform, systems, performance, QA, automation, and security engineering.

Work Terms

• Remote, hourly engagement.
• This listing is a standing pool and does not itself result in an acceptance or rejection decision.
• Project invitations are based on relevant stack or specialty. A specific project listing will identify its rate, hours, and client, and any hiring decision occurs for that project.
• Contact timing depends on available client projects and may range from a week to several months.

Compensation

Typical project rates range from $70 to $150 per hour, determined by project scope and depth.

Application Process and Eligibility

• Apply with a resume, confirm your work location, and complete a short AI interview of about 20 minutes.
• Your background will be verified before you are added to the software engineering talent pool.
• Completing the domain-expert interview and joining each network for which you qualify can improve matching opportunities.
• You register once to be eligible across all networks for which you qualify.