We build the Agentic OS for insurance case work. AI agents that draft objection rulings, summarize patient records, process claims, handle member support, the repetitive work that buries case workers 30–50% of their day. Live in days, not months. Getting smarter from every case through what we call the Context Brain & Cortex.

Customers are already seeing 80% faster processing and 95%+ quality rates. This is not a research project.

Here's what most AI companies get wrong: they think the hard part is the model. It's not. The hard part is building the interface where a skeptical insurance case worker — someone who's been doing this for 20 years, trusts an AI agent enough to delegate real work. That's a frontend problem. And it's the most important one we have.

What you'd build:

The surfaces where humans and AI agents collaborate on actual case work. Agent dashboards, real-time task delegation flows, document review interfaces, the Context Brain visualizations. You'd make autonomous systems feel legible, trustworthy, and fast, for users who have zero patience for anything that wastes their time.

What matters:

You're fast, you have taste, and you ship. Next.js/TypeScript. You've built complex, stateful UIs where things happen in real time and the user needs to feel in control. You think about UX like a product person, not just a pixel person. Bonus if you've thought seriously about human-AI interaction beyond chat bubbles.

What doesn't matter:

Your degree. Your years of experience. Whether you can whiteboard a red-black tree.

What you get:

Early team at a funded startup (Antler, AI.Fund, Earlybird) defining what enterprise AI looks like when it's not a demo. Equity. Munich-based, flexible on remote. A product that people actually use for real work, today.

How to apply:

• Send a resume. Send us 2–3 of the best UIs you've ever built links, screenshots, whatever shows the work. We care about what you've shipped, not what you've listed.