Skip to content
All open roles

AI engineer

  • AI
  • Mid-level
  • Remote
  • Full-time
  • 3+ years

Build AI assistants, agents and automation that run in production for our clients and inside our own products, with the testing and guardrails that make them dependable.

Apply for this role

Posted 18 October 2026

About the role

Our clients want AI that does a real job: answering customers, processing documents, keeping CRMs up to date. You will design and build those systems, from prompt and retrieval design to the APIs and interfaces around them.

We care about evidence. You will build the test sets that show an assistant works, measure it before and after every change, and keep cost and response times predictable.

What you will do

  • Build AI features and assistants using large language models and our clients' own data
  • Design retrieval pipelines over documents and databases, and keep answers grounded with sources
  • Build agents that call tools and APIs safely, with human approval for sensitive actions
  • Create evaluation sets and automated checks for answer quality, and track them over time
  • Monitor cost, latency and failure modes in production, and fix them
  • Automate business workflows by connecting AI to CRMs, help desks and back-office systems
  • Explain what AI can and cannot do to clients in plain language

What you will bring

  • 3 or more years of software engineering, with Python or TypeScript as a main language
  • Hands-on experience shipping at least one LLM-based feature to real users
  • Understanding of retrieval, embeddings and vector search
  • Experience building and consuming APIs, and working with relational databases
  • A habit of measuring results instead of trusting demos
  • Clear written English

Nice to have

None of these are required. If you have some, tell us about them.

  • Experience with agent frameworks such as LangGraph or the Vercel AI SDK
  • Running open models with vLLM or similar, including on-premise
  • Workflow automation with n8n or Temporal
  • Speech-to-text or voice assistants

Your first 90 days

  1. First 30 days

    Learn our AI toolkit and ship an improvement to an existing assistant, measured against its test set.

  2. By 60 days

    Lead the build of a new assistant or automation for a client.

  3. By 90 days

    Own the evaluation and monitoring for the systems you built, and propose what to build next.

What we offer

Remote-first
Work from wherever you do your best work, with at least four hours of overlap with your team.
Flexible hours
We care about outcomes and clear communication, not about when you are online.
Proper equipment
A modern laptop and the tools and licences you need from day one.
Learning budget
A yearly budget for courses, books, certifications and conferences.
Paid time off
Paid annual leave, public holidays and sick leave, and we expect you to use them.
Real ownership
Small teams, direct contact with clients, and work on our own products as well as client platforms.

How hiring works

  1. 01

    Apply

    Send a CV or a link to your work. A short note about what you want to do next helps.

  2. 02

    Intro call

    Thirty minutes with someone from the team you would join. Questions go both ways.

  3. 03

    Technical session

    A practical conversation about real work: code you have written, a system you have run, or a small take-home you choose.

  4. 04

    Offer

    A written offer with the details clear. We aim to finish the whole process within two weeks.

Other open roles