Founding AI Engineer

Location: Central London, hybrid (3 days in office)

Salary: Up to £85,000 + up to 5% equity

Reports to: Founder

An early-stage, well-funded AI company is looking for its founding-level AI Software Engineer. You'd be the senior AI hire, reporting directly to the founder, owning the AI layer of a Type Script-first product that's already being used by enterprise clients in a regulated industry.

Base is up to £85k, but the equity is up to 5%. For the right person, this is the kind of role you'll look back on as the one that mattered.

They build AI-native agentic software that automates complex operational workflows inside large, regulated organisations. The product replaces slow, manual processes with intelligent automation that learns from existing ways of working, escalates to humans when needed, and integrates into the client's existing systems rather than sitting alongside them.

The company has real paying pilots with enterprise clients, committed investment for the next round, and around 18 months of runway. They were selected from thousands of applicants for a prestigious industry accelerator programme earlier this year. This isn't a pre-revenue idea, it's a product that people are already using and paying for.

This isn't a role where AI sits in a separate service that the main product occasionally calls. You'll be integrating AI capability directly into the Type Script-first product itself, making architectural decisions about how it works, and owning the reliability of what you ship.

Specifically:

• Building and shipping AI features end-to-end inside a production Type Script/Next.js product
• Owning evals and accuracy measurement, designing how the team actually knows whether AI systems are performing reliably, not just whether they've deployed
• Deploying AI systems securely into production, including key management and the operational complexity that comes with enterprise-grade regulated environments
• Staying full-stack aware across the product, even where you're not building every screen, so AI capability integrates cleanly rather than creating bottlenecks
• Acting as a technical interface with enterprise clients on features you've built, explaining AI behaviour in plain language to non-technical stakeholders

The current stack is Next.js, Type Script, PostgreSQL, Docker, Kubernetes, and Lang Chain/Lang Graph.

Three things are non-negotiable, and they're tested directly in the interview:

• Type Script-first. This is a Type Script-first codebase, not a Python service with a Type Script wrapper. If your background is Python-heavy but you've genuinely made the shift to Type Script-first development, that's fine, but you'll need to demonstrate it clearly.
• Agentic systems in production. Not prototypes. Not tutorials. Things you've shipped, that people use, that you've had to keep running.
• Evals. You need to be able to answer the question "how do you actually know your AI system is performing reliably?" with something more rigorous than "we tested it before we deployed."
• Beyond those three, the cultural bar matters as much as the technical one. The team is small, every hire changes the shape of the company, and the founder is explicit about wanting someone who can operate independently, communicate clearly with clients, and break through problems without needing close management. If your instinct is to wait for a ticket, this probably isn't the right environment.

• FAANG-only or large enterprise-only backgrounds
• Python-first profiles without demonstrated Type Script depth
• Visa sponsorship required (this role cannot sponsor)
• Outside London or unwilling to commute to Central London 3 days a week

• Base salary: up to £85,000
• Equity: up to 5%
• The equity is a meaningful part of this offer, not a footnote. The company is pre-Series A with committed investors. For the right person, this is a real stake in something being built from close to the ground floor.