Skip to content

How we work

Deliberately
unremarkable.

Consultancy goes wrong in predictable ways: vague scope, hourly billing, a six-month discovery phase, a final report nobody reads. We have designed the engagement to make each of those difficult.

The engagement

Five phases, and you can stop after any of them

01

Discovery call

30 minutes · Free

You describe where the time goes. We ask the questions that separate an automation problem from a process problem, and tell you which one you have.

You end up with

A straight answer on whether this is worth pursuing, and roughly what it would cost.

02

Audit and roadmap

2–4 weeks · Fixed fee

We interview the people doing the work, map the real process, and score every opportunity on value, feasibility, verifiability and change friction.

You end up with

A prioritised roadmap with costs and payback, detailed enough for any competent team to build from.

03

Build

3–10 weeks · Fixed fee

Short cycles with something to look at every week. Built on your accounts and your infrastructure, with your team in the room, so nothing arrives as a black box.

You end up with

A working system in production, with monitoring, alerting, error handling and a run history you can inspect.

04

Handover

1 week · Included

Documentation written for the person who will maintain it, live training with whoever owns it, and a clear escalation path if something goes wrong at 2am.

You end up with

Full ownership: code in your repository, automations on your subscriptions, credentials in your hands.

05

Optional partnership

Monthly · Cancel anytime

Some clients keep us on for continuous improvement and incident cover. Plenty do not, and that is a success too.

You end up with

Agreed monthly capacity, maintenance of everything we built, and priority response.

The deal

Both sides of it

What we need from you

  • One decision-maker who can approve scope without a committee
  • Access to the people who actually do the work, not just their managers
  • Credentials and system access inside the first week
  • An hour a week for a demo and honest feedback

What you get from us

  • A fixed fee agreed in writing before anything starts
  • Something visible to review every single week
  • Direct access to the people building it, not an account manager
  • A clear answer when we think an idea is not worth your money

Engagement shapes

What a project looks like in practice

Three representative engagements, described by shape rather than by client.

These are illustrative. They describe the type and scale of work we take on, not named past clients. Real, attributed case studies will replace them as clients agree to be named.

Professional services firm · 40 staff

The problem

Every new enquiry was manually researched, logged in two systems and assigned in a weekly meeting. Response times ran to three days.

The approach

Intake automation: form to enrichment to scoring to CRM to owner assignment, with a drafted first reply queued for human approval.

Automation build · 4 weeks

B2B software company · 120 staff

The problem

Support specialists spent most of their day answering questions already documented somewhere in a decade of tickets and help articles.

The approach

Permission-aware retrieval layer over the help centre and ticket archive, then a support agent grounded in it that escalates with full context.

RAG + agent build · 9 weeks

Multi-site operator · 300 staff

The problem

A monthly reporting pack was assembled by hand from six systems, taking two people the better part of a week every month.

The approach

Automated collection, reconciliation and assembly with variance checks, delivered to the channel where the decisions are actually made.

Automation build · 5 weeks

Questions

Working with us

How quickly can you start?

Usually within two to three weeks of the discovery call. Audits can often start sooner. We deliberately keep a small number of concurrent engagements, so occasionally the answer is a month — we will tell you honestly rather than starting and stretching.

What if the build does not work?

We scope so that failure is visible early and cheap. Each build has an agreed success measure defined before we start, and weekly demos mean you see problems in week two, not week nine. If a system cannot hit its measure, we say so and we stop — we would rather lose the remaining fee than ship something you will not use.

Do you work with our existing team or agency?

Often, yes. We can run strategy and hand a specification to your developers, build alongside them, or take a defined piece end to end. We are not precious about who does what as long as ownership is clear.

Do you sign NDAs and DPAs?

Yes, as standard. We are happy to work under your paper. We can also work entirely within your infrastructure and accounts if your security posture requires it.

What sizes of business do you work with?

Typically 10 to 500 people. Below that the fixed costs of a proper engagement rarely pay back, and we will usually point you at tools you can run yourself. Above that you generally need a partner with more delivery capacity than we carry.

Next step

Let’s find the first thing worth automating.

A 30-minute call, no pitch deck. Tell us where the time goes and we’ll tell you honestly whether AI is the answer — and what it would take.

Free · 30 minutes · No obligation