AI Builder

Job description

**Role Summary** You will act as a chief of staff, directly working with the leaders. You will own the full loop: identify a high-impact business problem, conceive the AI-powered solution, build the production system, ship it, and own the outcome — as a single person. The problems change every few weeks as the company’s priorities shift. One month you’re building a multi-agent system for merchant risk detection or transaction success-rate optimization. Next, you’re deploying an internal operations copilot that processes hundreds of thousands of events a day and building products for the customers. What stays constant is the expectation that you close the loop end-to-end — from a messy business brief to a live system with metrics moving — in days, not quarters. **You Know Who You Are** You have deep expertise across the full building journey: - **Decompose business problems from first principles.** When a messy brief lands, you break it into data, models, integrations, automations, and human touchpoints — and make the right call on what to build vs. buy vs. defer. - **Design and build the full system yourself.** Architecture, backend services, data pipelines, LLM orchestration, frontend — you own the spec nobody writes for you. You write production code, not demos that die after a meeting. - **Ship to production and own the outcome.** Phased rollouts, alert thresholds, business metrics. When something breaks, you fix it. When numbers are below target, you know why. - **Switch altitude — technical depth to C-suite context.** Debug a model’s internals in the morning. Present the roadmap to leadership in the afternoon. Review legal terms for an AI partnership before the end of day. - **Multiply AI capability across the organisation.** You make teams across Sales, Marketing, Risk, and Engineering better at identifying and executing AI opportunities on their own — through frameworks, reference tools, and workshops that change behaviour. You’ve been using AI tools for over a year — Claude Code or Cursor is your default environment, not a novelty. You orchestrate agents, build with skills and MCPs. AI assistance is muscle memory. You’re a polyglot — picking up new stacks fast because AI handles the ramp. You thrive in ambiguity and default to action. **You will be a good fit if you bring** - **Founder mindset:** you’ve started something (a company, a product, a team) and taken it from zero to live with your own hands. You don’t wait for specs. - **Full-stack capability:** backend services, data pipelines (Kafka, Airflow), LLM orchestration (OpenAI, Anthropic, Gemini), frontends (React), and deployment infrastructure (GCP, Docker). You don’t wait for other teams to unblock you. - **LLM & agent depth:** hands-on with multi-agent systems, RAG, prompt engineering, and production AI pipelines — not just API wrappers. - **Product instinct:** you ask “why” before “how.” You can scope a business problem down to a shippable prototype and articulate business value to non-technical stakeholders. - **Speed & autonomy:** you can take a problem from discovery to demo in under two weeks. You move faster than the process and create structure only where it helps. - **Influence without authority:** you align stakeholders across Sales, Risk, Engineering, and Legal by being clear, credible, and fast — without having direct reports in any of those functions. **You will not be a good fit if you** You need detailed specs handed to you. You prefer deep specialisation over breadth. You’re uncomfortable with ambiguity. You optimise for code elegance over customer outcomes. You hand off after designing or after shipping. You treat “not my job” as a valid answer. You need a stable roadmap to do your best work. You still thrive living in a 2-week sprint.