Software Engineer - AI Agent Evaluation (Remote) at Toloka AI

aitrainer; New York, Northern (NY, KY); Hybrid; USD 57,000 - 81,000; Full time

Job summary

Mindrift connects specialists with project-based AI opportunities for leading tech companies. You will create challenging tasks and evaluation criteria within realistic simulated environments, forming believable development histories and ensuring tasks are solvable by AI agents.

The role emphasizes writing tests that accept multiple correct solutions and requires 5+ years in software development plus strong Python/JS stack experience.

Qualifications

• 5+ years in software development.
• Proficiency with Python (FastAPI) and JS/TS (React).
• Experience writing tests (functional, integration).
• English proficiency – B2+.
• Familiarity with Docker, Postgres, Kafka, Redis.

Responsibilities

• Build realistic developer environments – a virtual company with codebase, infrastructure, and context.
• Design tasks from intermediate states of these environments – craft prompts and success criteria.
• Write tests that verify agent solutions and accept all valid approaches.
• Iterate on tasks and tests based on QA feedback and refine until evaluation is fair and robust.

Skills

Python (FastAPI)

JavaScript/TypeScript (React)

Docker

Postgres

Kafka

Redis

Testing (functional, integration)

English proficiency (B2+)

Job description

Description Please submit your CV in English and indicate your level of English proficiency. Mindrift connects specialists with project-based AI opportunities for leading tech companies, focused on testing, evaluating, and improving AI systems. Participation is project-based, not permanent employment.

What this opportunity involves

You'll create challenging tasks and evaluation criteria within realistic simulated environments:

• Build realistic developer environments - a virtual company with codebase, infrastructure, and context (tickets, docs, conversations) that forms a believable development history
• Design tasks from intermediate states of these environments - craft the prompt, define what "solved" means, and ensure the task is solvable by an AI agent
• Write tests that verify agent solutions - accept all valid approaches and reject incorrect ones, neither too strict nor too lenient
• Iterate on tasks and tests based on QA feedback - review agent solutions, analyze failures, and refine until the evaluation is fair and robust

What this is NOT

• Not data labeling
• Not prompt engineering
• Not writing code from scratch - the agent writes most of the code; you guide and evaluate

What we look for

• 5+ years in software development
• Core stack: Python (FastAPI), JavaScript/TypeScript (React), Docker, Postgres, Kafka, Redis
• Experience writing tests (functional, integration)
• English proficiency - B2+

Why this is hard

Frontier models are already good at coding. Creating a task that genuinely challenges the best models is non-trivial. You need to deeply understand where models fail and what scenarios reveal the difference between a good and a bad solution. Tasks have many valid solutions - writing tests that accept all correct solutions and reject incorrect ones is harder than it sounds.

Compensation

Up to $50/hr equivalent , depending on level and pace. Tasks are estimated at ~20 hours each; you set your own schedule.