An expert inside your team, building AI into the work.
We don’t hand you a report and leave. An expert joins the people who produce your deliverables, learns how the work is really done, and builds AI into it with them, inside the tools you already use.
Most of the work is ours. Some of it can only come from you.
We can’t fit AI to work we haven’t seen. These are the things only your firm can provide.
- Time with the people who do the workEnough to watch real work being done and to test changes with them.
- Examples of real workRecent deliverables, templates and checklists.
- Access to your toolsAccess to the systems the work happens in, under rules you set.
- One decision-makerSomeone who can approve tools, data rules and changes.
- IT sign-offFor any AI subscriptions, which we list with their costs before you commit.
The way we work matters as much as the tools.
We work where the work happens
Next to your team, in your tools and on real client work, not from a slide deck.
We build in your environment
Your systems, your templates, your data rules. Nothing to migrate to.
We adjust as we go
Setups change as people use them. What doesn’t help is dropped.
We judge by approved work
Success is client-ready work your team approves, not a demo.
We hand over properly
Every setup is documented and owned by someone on your team before we step back.
You stay in control
Nothing reaches a client or goes live without your approval.
Five steps, with a clear exit after the third.
| Step | What happens | What you give | What you get |
|---|---|---|---|
| 1. Fit conversation | We talk through your services, your team and where the effort goes | A conversation | A straight answer on whether this is worth doing |
| 2. Join your team | An expert works alongside each role, in your systems, doing real work with them | Time with the people who do the work, sample work, tool access | A first-hand picture of how your deliverables are actually produced |
| 3. Choose the work | Together with your team, we weigh each task on volume, effort, predictability, risk and data sensitivity | A review with your decision-maker | The tasks worth changing, the expected effect, tools needed and their cost, and the tasks to leave alone |
| Decision point: stop here with what you have learned, or continue. No obligation either way. | |||
| 4. Build it into the work | Tools set up inside your systems; instructions and templates written with the people who use them; adjusted as real work comes through | People to test with, IT approval | Working setups inside your tools, refined on live work |
| 5. Hand over | Documentation, training for whoever owns each setup, and a comparison with where you started | An owner for each setup | A playbook per task and a before-and-after view of approved deliverables |
Everything we create is yours to keep and change.
A map of the work
Who produces what, and how each deliverable comes together.
A clear recommendation
The reasoning behind every choice, including the tasks we advise against.
Working setups
Inside the tools your team already uses, not a separate system to log into.
A playbook per task
What AI does, what a person checks, and what to do when it gets something wrong.
Written data rules
Which tools may see which information, approved by you.
A before-and-after view
Approved deliverables compared with where you started.
Fixed fees, agreed before any work starts.
We sell expert time and finished work, not software licences or subscriptions. If a task needs an AI tool, you pay its provider directly, and we list which ones and what they cost before you commit.
| Engagement | Covers | Pricing |
|---|---|---|
| Assessment | Steps 2 and 3 for one team or service line | Fixed fee, agreed before we start |
| Integration | Steps 4 and 5 for the tasks you chose | Fixed fee, agreed before we start |
| Ongoing support | New tasks, tool changes, new staff, keeping results from fading | Retainer, if you want it |
| Custom tools | Software for a process that has proved its value | Scoped separately |
We build software only when the numbers say so.
Sometimes one process is worth more than general tools can deliver, such as a high-volume document check or a report produced hundreds of times a year. We recommend a build only after the process has run with existing tools and the result has been measured.
Our background in product design means anything we build is designed around the people who will use it.
Someone who knows the tools and how to study work.
Every engagement is run by someone with deep knowledge of the AI tools and a background in user research and product design. That second skill matters most: the job is to watch how work actually gets done before changing it. Your team brings the expertise in your field; we bring the expertise in the tools.
Find out if your firm is eligible
A few questions about your firm and your work, and an evaluation based on research into how AI is used in your field.