Skip to content

Careers

Building Kai

A small team in the United Kingdom, building an AI agent that learns how work gets done and does it for you, personal to your company. Your work ships in the product from the first week.

See open roles ↓

Open roles

1 open

Nothing here fits? Tell us what you would do instead.

Send a general application

Engineering

Founding Engineer

Mirfield · Hybrid · Full-time · Posted 27 Sept 2026

GPU compute is one of the fastest growing asset classes in the world, and the capital behind it still moves on spreadsheets and email. Xamar AI is building the financial infrastructure for the GPU cloud economy: the layer lenders and compute operators will run on. As our Founding Engineer, you will build it from the first line of code. It will be AI native from day one, shipped fast, and trusted with real money.

What you'll own

  • Architecting and shipping our core platform end to end, from the first commit to production.
  • Designing agentic AI systems that do real financial work: reliable, auditable and built to scale.
  • Choosing the stack, setting the engineering bar, and shaping the team that grows around you.
  • Turning each month's new AI releases into a product advantage, fast.

What you'll bring

  • Deep, hands-on experience building and shipping AI/ML systems in production, not just prototypes.
  • Strong Python and Java, and modern React on the front end.
  • Real experience with agentic AI: LLM orchestration, tool use, retrieval, evaluation and guardrails.
  • A genuine obsession with the frontier. Whatever launched this month, you have already tried it.
  • Sharp problem solving: you take unclear, unsolved problems apart and ship the answer.
  • Clear communication and generous teamwork. You make the people around you better.

You'll thrive here if

  • You would rather build something new than maintain something old.
  • Startup pace, real ownership and a blank page energise you.
  • You want to solve problems nobody has solved yet.

What we offer

  • Meaningful founding equity. You will be well compensated and own a real part of what you build.
  • Hybrid working from our HQ in Mirfield, West Yorkshire.
  • Refreshments at HQ, and parking on site.
  • Regular team activities and some fun alongside the serious work. We take the problem seriously, not ourselves.
  • A full package of perks, growing as we do.

Why this, now

Why now.

  • The category is new.

    Companies run on scattered work. An agent that learns them, and keeps what it learns private, does not exist yet.

  • The work is concrete.

    A conflict caught at 2am. A slip flagged a week early. Everything we build is measured against moments like that.

  • The team is small.

    Small enough that what you own is visible in the product within weeks, and nobody is more than one step from the decision.

How we work

How we work.

  1. Do, don't display.Kai does the work, rather than reporting on it.
  2. AI reads. Code decides.Interpretation is a model's job. Every number is checked by real code.
  3. Private by default.What a company teaches Kai stays with that company. We build as if it were ours.
  4. Write it down.Decisions are recorded with their reasoning, so nobody re-litigates them six weeks later.
  5. Small team, whole problems.You own something end to end, including the parts that are not your speciality.

Equal opportunity

We hire on what you can do and how you think. We do not screen on where you studied or who you worked for, and we do not discriminate on any ground protected under the Equality Act 2010.

If you need an adjustment at any point in the process, tell us and we will make it. Asking has no bearing on the outcome.