Legion Health

Founding Engineer | 👨🏻‍💻 AI-Native Ops Infrastructure (YC S21, $2M+ ARR, $6.3M+ Raised)

San Francisco, US Full-time

Build AI-native care infrastructure. Ship into real patient care. Operate at speed.

Hey—I’m Daniel, co-founder & CTO of Legion Health (YC S21). We’re building the AI-native operations layer for psychiatric care: scheduling, intake, documentation, billing, risk detection, and coordination. Not “doctor-in-a-box”—the operational backend that mental health actually runs on.

We run our own psychiatric practice. The systems you build ship directly into real clinical workflows. Our agent infrastructure already supports 2,000+ patients with one human, and we’re hiring a Founding Engineer to own core backend + agent systems end-to-end.

You’ll architect our event-driven backend (Node.js / TypeScript / Supabase / AWS), build real LLM agents with tool use + memory + retries, shape internal operations UX that both humans and agents co-use, and ship product end-to-end. You don’t need to come in as a pure backend engineer—we’re open to strong product/full-stack engineers who think in systems and are excited to deepen their backend + AI infrastructure skills.

This role is in-person in San Francisco. Roles evolve fast here—including mine. I’m looking for someone who wants to grow into owning major parts of the system, not someone relying on me staying in a narrow CTO role forever.

🚀 If you’ve built complex systems or shipped product end-to-end with real ownership, think in events/state/flows, and want to deepen your backend + AI systems skills—we should talk.

🚀 Y Combinator Company Info

Y Combinator Batch: S21
Team Size: 11 employees
Industry: Healthcare -> Consumer Health and Wellness
Company Description: AI-native psychiatry, built for scale and covered by insurance

💰 Compensation

Salary Range: $130,000 - $180,000
Equity Range: 0.1% - 0.6%

📋 Job Details

Job Type: Full-time
Experience Level: 6+ years
Engineering Type: Backend
Time to Hire: 3

🛠️ Required Skills

Next.js Tailwind Machine learning Node.js React TypeScript Deep Learning Reinforcement learning (RL) Software Architecture PostgreSQL

🎯 Interview Process

We bias toward real signal over ceremony and move quickly.
  1. Intro call (20–30 min)—quick conversation with me (Daniel) about what you’ve built and what you want.
  2. Systems / portfolio deep dive (45–60 min)—walk through 1–2 systems you shipped; architecture, tradeoffs, failure modes.
  3. Practical work trial (1–2 hours async)—short, realistic backend/LLM-systems exercise. No leetcode, no puzzles.
  4. Final onsite (1–2 hours)—meet the team, pair on a real issue, discuss how you'd own a domain.

We typically go from first conversation → offer in 7–10 days.