
Mindrift
Expert-driven AI training platform by the team behind Toloka, connecting domain specialists with generative AI improvement projects.
- Open jobs
- 12
- Typical pay
- $69.75–$77.50/hr
- Rating
- 0.0
- Would recommend
- 0%
- Work type
- AI Training
- Company size
- 51-200
- Headquarters
- Newburyport, USA
About Mindrift
Mindrift is an AI training platform founded in 2024 by the team at Toloka, a data partner in the generative AI market with over a decade of experience. The platform connects domain experts, including professionals in medicine, finance, education, linguistics, biology, mathematics, computer science, and engineering, with projects aimed at making generative AI models more helpful, reliable, and responsible. Contributors work on tasks such as writing, reviewing, evaluating, coding, and refining AI outputs, all requiring domain-specific knowledge. Mindrift operates as a fully remote, freelance platform with flexible schedules and no location requirements. The platform offers project-based payments ranging from $15 to $100+ per hour depending on expertise, skills, and project needs, with payments processed bi-weekly. Mindrift has team members across the Netherlands, Australia, Indonesia, Malaysia, New Zealand, and the Philippines.
Mindrift pay by role
| Role | Pay |
|---|---|
| Writing / generalist | USD 15–30/hr |
| Data science / ML engineer | USD 40–90/hr |
| Senior Python engineer | USD 40–80/hr |
| Legal / medical consultant | USD 40–100/hr |
Advertised maximums from mindrift.ai job listings; earnings vary by region and task.
Who Mindrift hires
Qualifications vary by project and domain. Many projects require specialized knowledge in specific fields such as medicine, finance, data science, mathematics, or engineering. A strong command of English at C1 (advanced) level or above is typically required. Educational requirements range from a Bachelor's degree for some roles to a Master's or PhD for specialized positions, for example, Data Science AI Trainer roles require a Master's or PhD in a quantitative field with at least 5 years of hands-on experience. Technical skills such as Python programming, SQL, web scraping, and familiarity with GenAI technologies (LLMs, RAG, prompt engineering) may be necessary depending on the project. All candidates must demonstrate adaptability, attention to detail, and the ability to follow complex guidelines. A laptop and stable internet connection are essential. The application process involves submitting a CV, completing assessments, and verifying identity.
How the Mindrift hiring process works
- 1
Submit CV · about 1 days
Apply on mindrift.ai with your resume and domain information.
- 2
Assessments · about 5 days
Complete qualification tests that match your chosen track.
- 3
Profile verification · about 3 days
Mindrift verifies identity and credentials.
- 4
Onboarding · about 3 days
Review guidelines, project expectations and quality standards.
- 5
Project or talent pool · about 14 days
Join active AI projects or a pool for future matches.
Open Mindrift jobs
Mindrift FAQ
Who runs Mindrift?
Mindrift is an AI-training platform operated by Toloka (Toloka AI AG, Switzerland). CEO and founder: Olga Megorskaya.
Who can join?
Domain experts, professionals, and academics, 70% of the active 20,000-strong network hold Master's or PhD degrees, across 90+ domains and 20+ languages.
What kinds of projects are there?
AI Workflow Engineer, ML Engineer, Data Science, Python Experts (biology / chemistry / math / physics / statistics), Engineering (civil / electrical / mechanical), Design, Writing/Editing, Legal, and Supply Chain consulting.
What can I earn?
Advertised range $15–$100+/hr by role. Examples from listings: Data Science up to $90/hr, ML Engineer $90/hr, Computer Science Expert $80/hr, Legal Consultant $75/hr. Earnings are samples, not guarantees.
How much does Mindrift pay?
Across 4 open Mindrift listings with posted pay, the median is $70 per hour, and most pay between $69.75 and $77.50 per hour.
Is Mindrift hiring right now?
Yes. Mindrift has 12 open AI training jobs listed here, updated daily.