Open positions

Senior AI Engineer

  • Engineering
  • Remote, UK & international
  • Full-time

About otum

otum is the data and automation engineering team for companies without one. We build the data platforms, back-office automation, and software companies usually need an in-house team for. Senior engineers scope each system, build it, and stay accountable for it.

How we work

We are a senior team, working across the UK and internationally. Teams stay small, everyone works directly with the client, and every build is reviewed by a senior engineer before it reaches them.

We are stack agnostic, which means we bring the right engineer to each job rather than bending every engineer to the same tools. You work in what you are genuinely strong in.

We work remotely, with a few hours of daily overlap for the client calls. Some meetings are fixed and there is the occasional trip to a client site. The rest is flexible, and we make time for learning and for reviewing each other's work.

Role

You build AI workflows into a client's own stack that do the manual document and data work, reliably enough to trust with real decisions. You make the model's output verifiable, route the uncertain cases to a human, and hand over code the client owns. This is the AI side of otum.

Team

You join the AI side of otum, working on our AI Workflows builds. Teams are small and you work directly with the client's operations and domain people. A senior engineer reviews every build before it reaches the client, and you review others' work in turn.

What you'll work on

Build AI workflows that read the documents and data a client runs on, extract the fields into a schema, and drive the next action, matching, updating records, routing, and flagging risks.

Layer deterministic checks over the model's output, business rules and validation a client can inspect, so a guess that breaks the rules is caught by code rather than posted.

Route low-confidence cases to a human with a per-field confidence score, and turn their fixes into new test cases.

Practise eval-driven development, measuring field-level accuracy on a golden set of the client's real documents before anything goes live.

Build it to run inside the client's own cloud, so their documents never leave and nothing you process trains anyone's models, with an audit trail on every run.

Sit with the client to scope the first workflow and agree what good enough to trust actually means.

What we're looking for

Several years building applied AI or ML systems in production, not just demos, with LLMs or document-processing pipelines.

Experience with structured extraction, evaluation (golden sets, field-level precision and recall), and confidence-based routing.

Comfortable layering deterministic validation over probabilistic output, and building human-in-the-loop review into a workflow.

Familiar with running models in a client's own cloud, model-agnostic, and integrating through existing APIs.

Able to work directly with a client, and to be honest about where AI is and is not reliable.

Location

Remote. You can be based anywhere in the UK or internationally, as long as you keep a few hours of daily overlap with the team and the client. Some client calls are fixed, and there is the occasional trip to a client site.

Send your CV and we will come back to you shortly.

Engineering edge,
without hiring a team.

Numbers you can trust, operations that run themselves, all built into your stack and owned by you. Tell us where you want the edge, and we'll build it.

Have a project in mind?

  • Quick response

    You'll hear from an engineer the same working day, not a sales sequence.

  • Clear next steps

    A short call, then a plan, and we start with Proof Week or a full build.

  • Fixed fee first

    A fixed fee agreed before anything starts, for Proof Week or the full build.