Product Engineer - Bolter

Role overview

An early-stage AI agent platform is hiring a founding product engineer to own the full lifecycle of its product, from rapid prototyping through to production. The platform turns plain-language descriptions into agents that carry out real work end to end, retaining context, learning how each user works, and producing shareable outputs such as trackers, dashboards, and custom tools. This is a lean, validation-stage role where the engineer works directly with the GM and shapes the core technical decisions for what the product becomes.

Responsibilities

- Own the technical roadmap across three areas: the agent runtime that turns a description into a working, reliable agent; the workspace where users steer their agents and consume what those agents produce; and the reliability layer that makes agent output something users can trust in production. - Make architecture calls with real trade-offs across speed, reliability, cost, and build-versus-buy decisions, including where AI does the work and where humans steer. - Ship working prototypes quickly and harden what proves out, treating this as a validation environment where polish follows proof. - Use AI tools as a force multiplier alongside a small human core and a fleet of specialised agents that support implementation. - Help recruit and shape the early engineering team as the company scales beyond the founding group. - Stay close to the product and contribute strong opinions on UX, not only infrastructure, while partnering directly with the GM.

Requirements

- A track record of shipping production software, ideally as a founding or early engineer at a product-led startup. - Strong TypeScript across the full stack, with some Go where it fits the problem. - Depth in at least one of: LLM application design and production AI reliability, real-time data infrastructure, or building developer-facing products and internal tooling. - Strong product instincts that connect technical choices to user outcomes and business impact. - Comfort prototyping fast and loose to validate a feature, then shifting into steady-state mode to harden what is worth keeping, including tests, edge cases, and reliability work. - Genuine interest in AI messaging and productivity products, comfort with ambiguity where textbook answers do not exist, and clear written communication as a first-class skill.

Nice to have

- Experience building AI products, productivity tools, or chat applications where an agent or assistant does real work for a user. - Background building consumer-facing productivity or messaging applications at scale. - Familiarity with LLM evals, hallucination mitigation, or production AI reliability at scale. - Experience with real-time streaming architectures or event-driven systems. - A public portfolio such as open source contributions, papers, blog posts, or notable side projects. - Comfort operating across the stack, from backend services up to frontend UX.