Build everyday AGI. Trustworthy, consumer-grade agents that redefine human–AI collaboration for millions. Software shouldn’t wait for commands; it should partner with you, amplifying what you can do every single day.
We’re a stealth team of elite founders and AI researchers, with backgrounds spanning Stanford, OpenAI, and DeepMind We’re industry leaders in mobile and computer-use agents, bringing these capabilities to consumer scale.
Grounded in years of agent research, our AI is designed with trustworthiness and reliability as core pillars, not afterthoughts.
We are supported by tier-1 investors who funded the first generation of AI giants; now they’re backing us to build the next: everyday AGI. (Watch the demo)
If you see possibility where others see limits, read on.
We're looking for experienced data annotators to help train and align our universal AI agents. You'll be evaluating agent behavior across computer and mobile interfaces, providing the nuanced feedback that shapes how our models learn and improve.
This is not a mechanical task. Your judgment defines what "helpful," "safe," and "aligned" mean in practice. You'll work directly with ML researchers to refine guidelines, surface failure patterns, and build the datasets that drive our next breakthrough.
You'll focus on quality, consistency, and insight, ensuring every annotation moves us closer to agents people can trust.
Evaluate agent trajectories: Review and annotate agent behavior across desktop, web, and mobile environments — labeling actions, identifying failure modes, and assessing task completion quality.
Judge decision quality: Evaluate agent decision-making at each step: Was this the right action? Was it efficient? Did it align with user intent?
Provide structured feedback: Deliver qualitative feedback on agent behavior, including edge cases, reasoning errors, and UI misinterpretations.
Collaborate with researchers: Work directly with ML researchers to refine annotation guidelines, surface patterns in model failures, and inform training priorities.
Build benchmark datasets: Contribute to high-quality datasets that drive measurable performance improvements.
1+ years of experience annotating agent-based trajectories, reinforcement learning data, or similar sequential decision-making tasks
Strong understanding of what constitutes model quality — you can distinguish between a correct action, a suboptimal action, and a hallucinated one
Familiarity with model alignment concepts: helpfulness, harmlessness, honesty, and how annotation choices influence model behavior
Track record of working alongside researchers — you're comfortable with ambiguity, can propose annotation schema improvements, and understand how your work feeds into training pipelines
Excellent attention to detail and ability to maintain consistency across high volumes of data
Clear written communication skills for documenting edge cases and providing actionable feedback
Models learn from data. Data quality determines model quality. Your annotations are the ground truth.
You will directly shape how our agents behave — what they prioritize, how they reason, and whether users trust them. The patterns you identify and the feedback you provide will inform the next generation of training runs.
🏢 All in, in person — work moves faster face-to-face
🚀 Ship by default — speed and polish can coexist
🤝 One band, one sound — radical candor, zero politics
🏥 Competitive company-sponsored medical, dental, and vision insurance
✈️ Top-tier relocation and immigration support
Send us:
A link — or 60-second video — of something you built and why it mattered
Your resume or LinkedIn
Two sentences on the hardest challenge you've solved
Every exceptional candidate hears back within 48 hours.
If you see possibility where others see limits, we'd love to meet you.
Designing Everyday AGI
