Founding Engineer

About us → Building the AI operating system for GTM GTM tooling is broken → We’re building the foundations to enable the Ai agents that fix it. Revenue teams are buried in countless tools, tangled integrations, and years of tech debt → data scattered across ten+ systems, stale, permissioned, and full of half-truths typed into Salesforce on a Friday afternoon. We're building the foundation that finally enables AI agents for GTM. Not chatbots that summarise calls. Agents that do the actual work in one structure system → researching accounts, drafting outreach, updating the CRM, triaging inbound, enabling forecasts, tracking key metrics, suggesting changes in strategy. The hard part isn't the agent loop. It's the data and context layer underneath, and the correctness layer on top. We're currently in stealth, backed by some of the UK’s top pre-seed investors. If this problem space excites you we’d love to chat What you'll work on * The engine and DSL. Deterministic core: same inputs, same output, every time. You'll extend it to new rule shapes without breaking logic already running in production. * Data in, numbers out. Import of messy customer data, an append-only ledger so every output traces back to the rows that produced it, period processing, user-facing statements. * AI in the product. Authoring and adjustment assistants, schema inference from a sample CSV, and everything that happens when a model or provider misbehaves. * The product itself. TypeScript end to end: Next.js, tRPC, Postgres, AWS. Small team, so you see the customer problem, spec it, build it, ship it. What you bring Required * 5+ years building production systems, strong TypeScript. * Experience building against at minimum one of these problems: * A system where the numbers had to be right and explainable → payments, billing, payroll, ledgers, financial reporting. * A rules engine, DSL, or workflow engine that non-engineers configure. * Import pipelines that turned messy customer data into something trustworthy. * An LLM feature in production, including fallbacks, tool-call handling and failure UX. * Postgres directly, not only through an ORM. * You test the thing that matters and write PRs that explain *why* . Strong signal * Fintech, payroll, or billing background. Adjacent counts: years on a trading system and never touched Next.js beats knowing every React hook and never having a production number be wrong. * You use AI coding agents daily and have opinions on doing it well. Claude Code is part of our workflow. * Multi-tenant systems with row-level access control. How we hire Four stages, no whiteboard puzzles, a fast feedback process between stages 1. Founders call (45 min). What you've built, what you'd want to own, mutual fit. 2. Engineering lead Interview (60 min). Walk us through a system you built where correctness mattered. 3. Technical deep-dive (up to 2 hrs). You and the technical team build something end-to-end together. 4. References. We'd love to talk with technical people you've worked with. What you get * Top-of-market base + meaningful early-stage equity, 4-year vest, 1-year cliff. * Generous hardware budget (MacBook Pro + monitor + misc or equivalent). * AI-first tooling → from IDE to terminal. * Hybrid: WFH when you need to focus, in-office when it adds value.