The problemWhy this keeps hurting

Most organisations don't have an AI problem. They have a workflow problem AI might solve. The pressure to "do something with AI" produces tool sprawl, stalled pilots and demos that never reach daily work. The missing step is rarely the technology. It is knowing which of your workflows actually reward automation, and designing the process, including the humans in it, around that answer.

Who it's forDoes this sound like your organisation?

Teams feeling the "we should be using AI" pressure without a clear starting point. Leaders whose AI pilots impressed in the demo and died in the workflow. Operations teams carrying repetitive knowledge work (triage, drafting, summarising, data shuffling) that eats hours a machine should absorb.

How we workObserve · Decode · Redesign · Install

Observe

Discovery. We map how work and information actually move through your organisation: where hours go, what repeats, what your team quietly works around.

Decode

Diagnosis. We identify which workflows genuinely reward AI, which need process fixes instead, and where automation would just accelerate a broken step.

Redesign

Design. We design the target workflow (the AI component, the human checkpoints, the failure handling) and pick fit-for-purpose tooling, vendor-neutral.

Install

Implementation. We build it alongside your team, measure it against the baseline, and stay until the new workflow is the default, not the demo.

What you getDeliverables, not decks

A workflow and opportunity map: where your hours actually go, and which workflows reward automation
A prioritised backlog with honest effort/impact calls, including the workflows AI should not touch
Human-in-the-loop process designs: checkpoints, escalation, failure handling
Implemented, measured workflows: adoption in daily work, not a pilot in a sandbox

Engagement shapeHow it typically runs

Discovery

Typically 1–2 weeks. Workflow mapping, opportunity scan, and a prioritised recommendation you can act on with or without us.

Design

Typically 2–4 weeks. Target workflow design, tool selection, and a build plan with measures agreed up front.

Implementation

Scaled to the work. Build, integrate, train, measure until the workflow runs without us.

Durations are indicative and agreed per engagement. Pricing is discussed once we both understand the problem: no rate card theatre.

QuestionsAsked honestly, answered honestly

What if AI isn’t the answer?

Then we say so. Sometimes the honest recommendation is a process fix, a decision framework, or nothing at all. Discovery is designed to find that out cheaply, before you commit to a build.

Do we need technical people in-house?

No. We design workflows your existing team can run, and we handle the technical build. Where ongoing maintenance is needed, we set it up to be as boring as possible.

Which tools do you sell?

None. We are vendor-neutral. We pick fit-for-purpose tools for your workflow and budget, and we tell you when the cheap option is the right one.

What about our data?

Data handling is designed in from the start: what leaves your systems, what a model sees, and what gets retained are decided explicitly with you, not discovered later.

Where AI-assisted reasoning should live inside your organisation permanently, this work often feeds into AFOS, our organisational reasoning system, in development and available for discovery partnerships.

Discuss the problem.

Tell us what keeps repeating. We'll reply personally with an honest read, including if this service isn't the right fit.

Discuss a project All consulting services

Put AI where it earns its keep.