Software Engineer (Founding Team)

Hatch Talent; San Francisco (CA); On-site; USD 180,000 - 220,000; Full time

Benefits offered by this job

Equity 1-2%

Job summary

Hatch Talent in San Francisco partners with an AI startup building multimodal agents to automate architecture and housing construction. We seek a Founding AI Engineer who blends strong software engineering with applied ML/AI experimentation, owning features from ideation to deployment.

You’ll work across agentic AI and multimodal LLMs, Python, LLM evals, and model distillation. Ideal candidates have 1–6 years in software engineering / ML, enjoy fast iteration, and already use Claude Code or

Qualifications

• 1–6 years in software engineering or ML
• Hands-on experience building AI/LLM applications
• Autonomous and able to run your own experiments
• Experience with Claude Code or Cursor is a plus

Responsibilities

• Build AI/ML experiments across the agentic AI stack
• Develop and test ideas quickly from concept to production
• Collaborate on tooling to accelerate AI experimentation

Skills

Python

ML Engineering

LLM Applications

Autonomy

Experimentation

SWE Speed

Tools

Claude Code

Cursor

Job description

$180k-$220k + 1–2% equity

Want to join an AI startup early enough that “founding engineer” actually means something?

We're working with an AI startup building multimodal agents that are tackling a pretty huge real-world problem: automating architecture and helping get housing built faster.

Their technology uses AI agents to understand architectural blueprints, assess building-code compliance and ultimately automate parts of the architecture process.

They’re now looking for a Founding AI Engineer to join the team in San Francisco.

This isn't a pure research role. They're looking for someone who combines strong software engineering with applied ML/AI experimentation ; this is ideal for someone who likes building, testing ideas quickly and figuring things out without needing a detailed roadmap.

You’ll be working across:

• Agentic AI and multimodal LLMs
• Python
• LLM evals and experimentation
• Supervised fine-tuning and model distillation
• Retrieval and context engineering
• Internal tooling for faster AI experimentation
• Real product problems alongside users

What they’re looking for:

1–6 years across software engineering / ML, ideally with hands-on experience building AI or LLM applications.

More importantly, you’ll be highly autonomous, comfortable running your own experiments and able to combine SWE speed with ML rigor. You should also be someone who already uses tools like Claude Code, Cursor or similar as part of your day-to-day development workflow.

If you're looking for a role where you can have a massive amount of ownership, work directly on agentic AI and join while the technical foundations are still being built, I'd love to hear from you.

Founding Engineer | Agentic AI | San Francisco | AI Engineer