Join the data engine powering frontier AI Labs.
We're a growing team working on meaningful problems. Our training data directly shapes how frontier AI models learn, reason, and interact with humans.
The work looks like this.
Top of market rates
Competitive pay calibrated to the depth of the work and your reliability across projects.
Flexible hours
Set your own schedule. Minimum 10-20 hrs/week, scale up to full-time as projects fit.
Remote-first
Work from anywhere. Async-friendly project management, weekly research syncs.
Real impact
Your work directly shapes how frontier models learn, reason, and interact with humans.
Open roles.
All roles are remote, full-time or part-time, with a minimum of 10–20 hrs/week.
LLM Data Trainer
Full-time / Part-time
Join our team creating high-quality training data for frontier AI models. You'll work on diverse tasks including RLHF, instruction tuning, and preference labeling to improve LLM capabilities.
- ─Perform RLHF (Reinforcement Learning from Human Feedback) labeling
- ─Create and curate training examples for instruction-following tasks
- ─Evaluate and rank model outputs for preference optimization
- ─Write detailed prompts and high-quality responses across domains
- ─Document edge cases and failure modes in model behavior
- ─Strong analytical and problem-solving skills
- ─Strong written communication skills in English
- ─Familiarity with LLMs and prompt engineering concepts
- ─Attention to detail and ability to follow annotation guidelines
- ─Experience with data labeling or content moderation (preferred)
Code Data Specialist
Full-time / Part-time
Help improve AI coding assistants by creating high-quality code training data. You'll write, review, and annotate code examples across multiple programming languages.
- ─Write clean, well-documented code solutions for diverse programming tasks
- ─Annotate code with explanations of logic and best practices
- ─Develop debugging and code review training examples
- ─Create test cases and edge case scenarios for code evaluation
- ─Review and rank AI-generated code for correctness and quality
- ─Strong understanding of algorithms, data structures, and design patterns
- ─Proficiency in Python, JavaScript, and at least one other language
- ─Excellent written communication for code documentation
- ─Experience with code review and software testing
- ─Understanding of LLM capabilities and limitations
Domain Expert Annotator (STEM)
Full-time / Part-time
We're seeking subject matter experts in STEM fields to create and validate specialized training data for LLMs. Help train the next generation of AI models in your area of expertise.
- ─Create expert-level Q&A pairs in your domain (Math, Physics, Chemistry, Biology, CS)
- ─Develop multi-step problem-solving examples with detailed explanations
- ─Review and improve existing training data for technical accuracy
- ─Validate model outputs for factual accuracy and reasoning quality
- ─Collaborate with the team to define quality standards for your domain
- ─Graduate degree (MS/PhD) in a STEM field or equivalent expertise
- ─Strong analytical skills and attention to nuance
- ─Ability to explain complex concepts clearly and accurately
- ─Ability to work independently and meet quality standards
- ─Experience in teaching, tutoring, or technical writing (preferred)
Safety & Alignment Specialist
Full-time / Part-time
Work on the critical task of making AI systems safe and aligned with human values. You'll create red-teaming datasets and help define safety boundaries for LLMs.
- ─Develop adversarial prompts to test model safety boundaries
- ─Identify and flag problematic or unsafe model outputs
- ─Evaluate model responses for harmful or biased content
- ─Collaborate with researchers on alignment methodologies
- ─Create training data for refusal and safety behaviors
- ─Background in AI safety, ethics, or related field
- ─Ability to think critically about potential misuse scenarios
- ─Strong written communication skills in English
- ─Attention to detail and ability to follow annotation guidelines
- ─Experience with content moderation or trust & safety (preferred)
We're always looking for talent
Don't see a role that fits? Send us your background and what you're excited to work on.