Founding AI Engineer

About the Company

A seed-stage company (founded 2025) on the frontier of agentic AI for the built world, building the technical leader in AI that checks whether construction blueprints are buildable and safe. Its multimodal agents review blueprints for building-code compliance, becoming domain experts on things like ADA and Wildland-Urban Interface rules — work that directly affects California's built environment within a new hire's first month.

The company grew out of a pro-housing organization in San Francisco, and its mission is to solve the housing crisis: if it wins, the cost of building housing drops meaningfully and rents get more affordable. It's one of very few places in tech where an engineer can work directly on housing, and that mission — shared by its investors — is the single strongest draw. It has raised $4.3M, is the technical leader in its space, and its product demo reliably gets strong engineers excited.

About the Role

A deeply self-directed founding engineering role at a tiny company (three people, moving to two as the CTO transitions out). The bulk of the work is running experiments end to end on the company's eval framework: generating hypotheses, implementing experiments, and analyzing results to improve the accuracy of multimodal agents that check blueprints for code compliance. Demands someone independent and self-unblocking with strong research taste.

Two profiles have worked here: an exceptional software engineer who wants to break into AI research, and an AI/ML engineer already established in the space. The work is heads-down and technical, with only ~2 hours/week talking to users (not a forward-deployed role). Expectations are high: roughly 55 hours/week with weekend availability, in person in SF every day, in exchange for a large chunk of ownership.

What You'll Own

• Improving the accuracy of the multimodal blueprint-compliance agents — building autoresearch, improving evals, running sweeps on experimental features you design
• The full experiment loop: ideas, implementation, and rigorous analysis on the eval framework
• Becoming a domain expert on the parts of the building code that matter (ADA, Wildland-Urban Interfaces, structural and fire safety)
• Wearing hats as needed: computer-vision experiments, lightweight data engineering, internal tooling, webapp improvements
• Weekly user contact and product dogfooding (~2 hours/week)

Must-Have

1. Independent and self-unblocking, able to direct your own work with minimal oversight
2. High velocity — fast at prototyping and shipping experiments
3. Rigorous about evaluation, with genuine research taste to steer high-risk, high-reward work
4. Either a strong SWE eager to grow into ML/AI, or an ML/AI engineer already fast at prototyping
5. Genuine mission alignment with solving the housing crisis
6. Able to work in person in San Francisco, ~55 hours/week with weekend availability

Nice-to-Have

• Background at a self-driving car company (e.g., Waymo, Zoox), where evaluation rigor is life-or-death
• Experience at highly regulated application-layer AI startups (medical, legal)
• Experience at climate-tech startups
• Strong academic background
• The curiosity of a policy wonk to digest building codes and architectural diagrams

Details

Employment: Full-Time  |  Location: San Francisco, CA — In-person (daily)  |  Experience: 2–8 years
Compensation: $160,000–$200,000 base + 0.5–2% equity. Free lunch/dinner, fully paid health with $1,000+/yr employer HSA, commuter ($100/mo) and wellness ($50/mo) benefits. No visa sponsorship.