About the RoleThis is a founding engineering role at a fast-growing, seed-stage startup building AI-powered autonomous product demo agents for B2B GTM teams. The company has tripled in the past month, is backed by top-tier investors (Series A/B range), and is now assembling a small A-team to scale. You will be one of two founding engineers, working directly with the CTO - setting the technical culture and architecture for everyone who comes after.
The technical surface area is broad and genuinely novel: voice, browser agents, LLM orchestration, generative UI, reinforcement learning, and self-improvement loops. There is no narrow lane here. The best problems are unsolved because nobody has had to solve them before.
No visa sponsorship available. On-site in San Francisco, 5 days/week required.What You'll DoYou'll own work across four core technical pillars:
- The Agent: LLM orchestration, conversation flow, tool use, and voice. Building agents that understand products they've never seen and explain them clearly to prospects with varying levels of context.
- The Experience: Embeddable, generative UI that adapts in real-time - fast loads, complex interactive demos, responsive across products and screen sizes.
- The Platform: Customer-facing tooling for configuring agents, managing conversation flows, reviewing transcripts, and measuring performance.
- Evals & Self-Improvement: The pipeline that measures agent quality, detects regressions, and feeds learnings back so demos improve without manual intervention.
- Infrastructure: Zero-cold-start reliability - when a prospect clicks "start demo," the agent responds instantly.
In practice, your week might look like:
- Diagnosing a conversation log to find where an LLM lost the thread during a pricing objection, then shipping a fix to the orchestration layer by lunch.
- Prototyping a generative UI widget that adapts its layout in real-time based on prospect intent - demo-able by end of day.
- Tracing an eval regression to a prompt change, rolling it back, and adding a test to prevent recurrence.
- Designing agent versioning infrastructure from scratch so customers can preview changes before they go live.
What We're Looking ForRequired:- 5+ years of full-stack engineering experience, having built and shipped products at a VC-backed B2B SaaS startup with paying customers.
- Proficiency in React (frontend) and Python (backend).
- Hands-on experience building and shipping production AI agent features: LLM agents, RAG pipelines, or agent orchestration frameworks (e.g., LangChain, LlamaIndex).
- Ability to design, implement, and maintain an evaluation pipeline for model outputs, including regression tracing and rollback mechanisms.
- Experience building versioning and preview mechanisms for deployed AI agents.
- Strong front-end skills and UX sensibility; demonstrated ability to build polished, customer-facing products.
- Comfort translating user feedback and product requirements directly into implemented features.
- Strong engineering judgment - you know when something is ready to ship and when it isn't.
- Bachelor's degree in Computer Science or a related technical field.
- Ability to work on-site in San Francisco 5 days per week.
Nice to Have:- Experience at a VC-backed sales technology or GTM tooling startup.
- Familiarity with the interactive demo / product experience tooling space (e.g., companies like Walnut, Demostack, Navattic, Arcade, Storylane, or similar).
Compensation & Benefits- Salary: $180,000 - $230,000 USD annually
- Founding engineer equity commensurate with role and stage
- Full-time position
Location- San Francisco, California, United States
- On-site, 5 days per week
- No visa sponsorship available
