The role
- COMP
- $180K - $230K
- EQUITY
- Competitive equity
- LOCATION
- San Francisco
- WORKPLACE
- On-site
- EXPERIENCE
- 1 - 5 years
- VISA
- None
- STACK
- React, NodeJS, Python, CSS, Langchain
- INDUSTRY
- Software Development, AI, B2B
/roles — ROLE_396
Founding Engineer (Full Stack)
Fast-growing startup whose AI agents give prospects instant, hands-on product demos and keep improving over time
The company
An early-stage AI company building autonomous agents that handle customer-facing work like live product demos, onboarding, and support, learning from every conversation.
- STAGE
- Series A-stage
- FUNDING
- $6M raised
- TEAM
- founding team of <10
- FOUNDED
- 2025
JD — the work
About the role
Instead of booking a call and sitting through a pitch, a prospect talks to the company's agent and gets a live, sandboxed demo within seconds. Demand is climbing fast after a recent seed raise, and the technical scope is wide: it spans voice, agents that drive browsers, reinforcement learning, and loops that let the system improve itself. Engineering today is the CTO plus contractors, and you would be one of two founding engineers defining how the team builds for every later hire.
What you'll do
- Trace why an agent fumbled an unfamiliar sales objection through its logs, then patch the orchestration code the same day
- Prototype generative interfaces whose layout adapts live to each prospect's interests, often ready to show the same afternoon
- Chase eval regressions to their cause, roll back bad changes, and add tests so they do not return
- Design versioning for agents from scratch, letting customers test updates safely before release
- Contribute across every layer: LLM orchestration and voice, embeddable UI, the configuration platform, evals, and instant-start infrastructure
- Handle the unglamorous parts too, including CSS fixes, migration scripts, and customer bug reports
What they're looking for
- Experience building systems real users rely on, with firsthand stories of how they broke
- Sound judgment about when something can ship, and the ability to move fast without eroding the system
- Genuine interest in closing the gap between LLM demos and production, using evals and careful design
- Willingness to go wherever the hardest problem is, whether frontend, backend, or infrastructure
- Proof you can move fast under pressure and also pull back work that is not landing
- Initiative to fix broken or unclear things without waiting to be told