Transform the AI landscape with Mindrift as an Evaluation Engineer. Bring your software development expertise to create challenging tasks that evaluate real-world AI coding agents. Mindrift focuses on connecting specialists with AI project-based opportunities.

As an Evaluation Engineer, you will craft evaluation criteria and realistic tasks for assessing AI coding performance. This part-time, remote project allows you to influence future AI model capabilities while fitting around your qualified commitments. Key Responsibilities:

- Create challenging tasks for evaluating AI coding agents
- Develop evaluation criteria in simulated environments
- Design tests to verify AI agent solutions
- Refine tasks based on quality assurance feedback
- Analyze failures to enhance evaluation fairness Requirements:
- Minimum 5 years in software development
- Core skills: Python, JavaScript/TypeScript, Docker
- Experience writing functional and integration tests
- English proficiency at B2+ level
- Master's or relevant Bachelor's degree required Utilize your software skills to support Mindrift's mission in AI evaluation.

📌 AI Evaluation Engineer at Mindrift (Toronto)
🏢 Mindrift
📍 Toronto