AI Engineer

Full-time·San Francisco, CA·1700 Montgomery St

About

We’re building Jenny, a home assistant. Jenny helps homeowners take a project from first idea to final payment — understanding scope, navigating bids, and keeping everything on track without the chaos that usually comes with it. We’re a small team in San Francisco working on a problem that touches nearly every household.

The Role

We’re hiring an AI Engineer to own the agentic systems at the core of Jenny. This is not a research role — you’ll design, build, and ship the agent workflows that homeowners and contractors depend on every day.

You’ll work at the intersection of LLM systems and product, making real architectural decisions about how agents reason, plan, use tools, and hand off between each other. The problems are hard: multi-step workflows, ambiguous user intent, long-horizon task management, and accuracy requirements that matter when real money is on the line.

What You’ll Work On

  • Design and own end-to-end agentic workflows — from user intent to structured output across multiple steps and tool calls
  • Build the planning and reasoning layer that allows Jenny to scope projects, clarify ambiguity, and drive tasks to completion
  • Develop tool-use and memory systems that give agents persistent context across a project’s lifecycle
  • Build eval frameworks to measure agent accuracy, reliability, and regression across workflows
  • Integrate external data sources (cost databases, permit records, contractor profiles) as grounded context for agent decisions
  • Collaborate closely with product to translate user needs into agent behavior — and push back when the right solution is simpler

What We’re Looking For

  • 3+ years of software engineering experience, with at least 1 year focused on LLM or agent systems in production
  • Deep understanding of agentic architectures — tool use, planning loops, multi-agent coordination, retrieval, and memory
  • Strong Python engineering skills; comfort working across the stack when needed
  • Experience building and running evals — not just shipping prompts, but measuring whether they work
  • High agency: you identify problems, propose solutions, and drive them to done without waiting to be told
  • Familiarity with frameworks like LangGraph, LlamaIndex, or similar is a plus — but we care more about first-principles thinking than framework knowledge

Details

  • Location: In-person, San Francisco — 1700 Montgomery St
  • Type: Full-time
  • Compensation: Competitive salary + equity

Apply

Send a short note about what you’ve built and why this problem interests you to jobs@hijenny.ai. No cover letter required — just tell us something real.