Open Position
Build and ship LLM-powered agents and workflows that help Trust & Safety teams make faster, more consistent decisions.
About Precognition Labs
Precognition Labs builds enterprise AI for marketplaces to build trust, improve safety, and stop abuse and fraud. Trust & Safety teams already have policies, tools, and experienced investigators. We add an AI layer that works inside those existing systems: it gathers the right evidence, reasons over policy, and helps people make faster, more consistent, and defensible decisions. We are a small, founder-led team from Snap, Airbnb, Google, and Amazon, backed by Antler.
The role
As a Mid-Level AI Engineer, you will turn large language models into dependable product. You will design the agents that read a case the way an expert investigator does, pull evidence from the tools our customers already use, reason over their policies, and recommend a decision with a clear explanation. You will own features end to end, from prompt and tool design through evaluation, deployment, and monitoring in production.
What you’ll do
- Design, build, and ship LLM agents and multi-step workflows for investigation, policy reasoning, and case summarization.
- Build tool integrations that let agents gather evidence safely from customer systems and the browser.
- Create evaluation sets and automated evals that measure accuracy, consistency, and policy adherence, and use them to guide every change.
- Improve latency, cost, and reliability through prompt design, caching, model selection, and fallbacks.
- Instrument agents in production, investigate failures, and close the loop with customers and specialists.
- Work closely with the founders and customers to turn real Trust & Safety workflows into product.
What we’re looking for
- 2–5 years of software engineering experience, with at least 1 year building products on top of LLMs.
- Strong Python and/or TypeScript, and experience shipping production services.
- Hands-on experience with prompt engineering, tool use and function calling, retrieval, and structured outputs.
- A habit of measuring: you build evals before you trust a change.
- Clear written communication and comfort with ambiguity in an early-stage company.
Nice to have
- Experience in Trust & Safety, fraud, risk, integrity, or content moderation.
- Experience with browser automation, Chrome extensions, or agent frameworks.
- Experience with Google Cloud, Cloud Run, or similar serverless platforms.
- Prior startup or founding-engineer experience.
Why join us
- Work directly with founders who have led Trust & Safety, integrity, and AI/ML at Snap, Airbnb, Amazon, and Google.
- Ship to real customers early. Your work will be in front of Trust & Safety teams within weeks, not quarters.
- Meaningful ownership and early-employee equity.
- Remote-friendly, with a bias toward clear writing and fast decisions.
- A mission that matters: making online marketplaces safer for the people who use them.
Interested?
Tell us about yourself and what you’d build. Apply on Wellfound and we’ll be in touch.
Apply on Wellfound