Founding AI Engineer
San Francisco, CA | Onsite | Full-Time
We're seeking a highly autonomous Founding AI Engineer to join a seed-stage agentic AI company applying multimodal AI to construction and architecture. Its agents analyze blueprints to determine whether buildings are safe, buildable, and compliant with regulations such as ADA, Wildland-Urban Interface, structural, and fire-safety requirements.
This is a highly autonomous founding engineering role focused primarily on improving the accuracy of multimodal agents.
The core workflow is:
HYPOTHESIS → EXPERIMENT → IMPLEMENT → EVALUATE → ANALYZE → ITERATE
Most of the engineer's time will be spent designing and running experiments against the evaluation framework rather than interacting with customers. The role can fit either a strong software engineer moving into AI research or an established AI/ML engineer who is also a fast builder.
What You'll Own:
- Multimodal-agent accuracy
- AI evaluation frameworks
- Autoresearch
- Experimental feature development
- Experiment sweeps
- Computer-vision experimentation
- Evaluation analysis
- Lightweight data engineering
- Internal tooling
- Occasional web application improvements
- Building-code domain knowledge
- Product dogfooding
- Limited weekly user interaction
