← नौकरियों पर वापस जाएँ
Code Evaluation
$100 - $200 प्रति घंटा

Senior Software Engineer, AI Training - US

अभी आवेदन करें → →

विवरण

Evaluate how

AI coding agents work with developers Review complete interactions with coding agents such as OpenAI Codex and Claude Code. The focus is whether the interaction shows the judgment an experienced engineer would expect: useful responses, helpful explanations and appropriate guidance. The assignment involves evaluation rather than production development. You will consider how an agent reasons and communicates, as well as whether its output is broadly correct.

What you will assess

  • Review coding interactions from beginning to end for usefulness and overall correctness.
  • Judge whether preambles, explanations and reasoning help the developer.
  • Assess whether the response reflects strong engineering judgment and builds trust.
  • Distinguish levels of response quality and explain why one response deserves a different rating from another.
  • Write clear, direct feedback on what worked, what failed and what felt misleading.
  • Help define strong interactions with tools such as Cursor, Codex and Claude Code. You should be comfortable making subjective but rigorous judgments. The work calls for explaining whether an answer gives helpful guidance and sounds like something a capable engineer would actually say.

Required experience

  • Senior, Staff or Principal engineering experience, or equivalent experience.
  • A strong background in TypeScript, JavaScript or Python.
  • Practical experience with at least one of OpenAI Codex, Claude Code or Cursor.
  • Deep familiarity with modern development workflows that incorporate AI assistance.
  • Ability to evaluate code without executing it or examining every line.
  • Written and spoken English at B2 level or above. Detailed written feedback and short recorded video explanations are required. Prior exposure to prompt design or evaluation workflows is helpful, as is experience mentoring senior engineers or defining engineering standards.

Pay, schedule and selection

This is a worldwide remote contract paying $100 to $200 per hour, depending on experience and location. The schedule is flexible, with 10 to 20 hours of work per week. The engagement is ongoing, with individual projects lasting from around two weeks to a few months. Evaluators who perform well may receive additional projects. Selection consists of one evaluation exercise completed at home with a recorded Loom walkthrough; there is no interview. Work can begin after you pass that exercise and an open project seat becomes available.

Details

नियुक्ति करने वाली कंपनी

G2i

Category

Code Evaluation

Location

United States

Languages Required

🇺🇸 English

Posted

Senior Software Engineer, AI Training - US
$100 - $200 प्रति घंटा
अभी आवेदन करें → →