Founding Engineer: AI-Native SaaS Platform (PropTech)
£55,000–£85,000 + equity in the future product entity | Hybrid, London | Full time (contract-to-perm)
THE SHORT VERSION
We're Beaumont & Partners (www.beaupartners.com) a fast growth, vertically integrated UK property investment and development business that has spent the last year building our own operating platform, a full multi-tenant SaaS product covering project delivery, procurement, finances, compliance and reporting. It already runs on a modern stack (Next.js, Tailwind, Supabase/Postgres with row-level security, Drizzle, Vercel, GitHub Actions) and is built almost entirely through agentic AI development: designs are prototyped in Claude Design, builds run through Claude Code against a strict rules file, and safety is enforced by automated CI gates including tenant-isolation checks on every pull request.
Phase 1 is replacing our internal tooling, nearly done, and the business will run on it day to day. Phase 2 is launching it as a commercial SaaS product to other property businesses. We're hiring the engineer who will own that journey.
THE ROLE
You'll be our first engineer, taking over the technical function from the founders so they can get back to growing the business. That means:
- Owning the build: directing AI-agent builds, reviewing every line that merges, owning merge decisions and releases
- Owning the safety net: the row-level-security model, the tenant-isolation CI gates, auth, you are the human who answers for data separation
- Improving the machine: we've refined an unusual and effective AI-native workflow; we want you to make it better, not throw it out
- Getting the product commercial-grade for external customers: security posture, reliability, the things a customer's IT team asks about before signing
- Working directly with the two founders, who own product design and will keep feeding you designs and testing features, but the technical decisions are yours
HOW WE BUILD (READ THIS BIT CAREFULLY)
This is not a conventional dev role. Almost all code here is written by AI agents working from design exports and a detailed rules file, with humans steering, reviewing and verifying. If you love hand-crafting every line, you will hate this job. If you already ship production software primarily by directing AI tools and you can critically review what they produce, catch the security mistake, and defend every line that merges, this is one of the most advanced setups you'll find, and you'll have full ownership of it.
YOU
- 5–8 years building production web software, with real experience shipping and running a SaaS product (multi-tenancy, auth, and Postgres experience strongly preferred)
- You genuinely work AI-natively today, agentic coding tools are your daily workflow, not a CV keyword
- You can review AI-generated code critically, especially around security and data isolation
- You're comfortable as a one-person engineering team with high autonomy, communicating in plain English with non-technical founders
- You want ownership: equity in the future commercial entity is part of the package, with the path to technical lead of a funded, revenue-backed SaaS launch
HOW THE HIRING WORKS
We do it properly: a conversation with the founders, a short technical exchange, then a paid contract trial (6–8 weeks, real work on the real product) before we both commit to permanent. You'll ship a real feature and audit our security setup, you'll know exactly what you're joining, and we'll know exactly who we're hiring.
TO APPLY
Message us with your CV and a short note answering one question: describe the most substantial thing you've shipped where AI agents wrote most of the code, what you built, how you directed the AI, and what you personally caught or fixed that the AI got wrong.
No agencies please.
Pay: From £65,000.00 per year
Work Location: Hybrid remote in London
