About this role
Shape how AI plans, reasons, and acts. Join one of our flexible, paid, remote projects now!
What's agentic projects
Agentic AI is the next frontier: models that don’t just respond, but decide and act. Your contribution has lasting impact: every scenario, test, or tool you create influences how autonomous AI agents reason, make choices, and interact with people.
- Shape how future AI systems behave in the real world
- Learn and use cutting-edge evaluation methods and tools
- Collaborate with a global team on an innovative project defining the standards for safe, reliable AI
What you get
Flexible remote work
Hands-on AI experience
Shape how AI learns
Global expert network
Current opportunities
Agent Evaluation Engineer
Validate and refine test scenarios, improve coverage and logic
Apply
What's Mindrift
Contribute to advanced agent evaluation with real-world impact. Work on your schedule, from anywhere. Paid, scoped, project-based work with no open-ended commitments. Join a global community of experts, QAs, and writers shaping the future of intelligent systems.
Toloka, the company that created the Mindrift platform, works with the following clients:
Toloka Platforms Terms of Use
Eligibility and Geographic Restrictions
FAQ
Code of Conduct
Help Center
Facebook ↗
LinkedIn ↗
Reddit ↗
All projects
How it works
Blog
Community
About us
Log in
Backed by Toloka with rates shown before you start, and tracks for many skill sets. Domains go quiet for stretches, and reviews are mixed on availability and support.
AITrainerGigs aggregates this listing from Mindrift.