For teams of 5 to 50
We come in, look at how the team actually works, and rebuild the bits where AI should be doing the lifting. The training sticks because we build it into the job, not into a slide deck.
What we usually find
Half the team is quietly using ChatGPT for everything. The other half is waiting for IT to bless something. Neither group is producing better work. That is not a people problem. It is a missing instruction problem.
The confident ones are winging it. The careful ones are stuck. Nobody has been shown what a good AI brief looks like, what to double-check, or when to walk away from the output. We teach the practice the way you'd teach a new hire. By doing the actual work.
Most workflows were built for a team without AI, then patched. The savings keep leaking out at the handovers. We rewrite the few workflows that matter and mark the spots where AI does the work, where it helps a person, and where a person stays in charge.
You are paying for tools nobody owns and a few internal builds nobody can maintain. We trim the kit down, give each tool an owner, and teach the team to build the small internal stuff themselves instead of waiting six weeks for it.
Who is doing the work
Sarah. Twenty years inside service businesses, doing the unglamorous half. The workflows, the SOPs, the why-doesn't-this-thing-talk-to-that-thing. Now with AI bolted to the side of it.
Service Design
Workflows the team can actually run on a Tuesday.
Applied Learning
Training built for adults with full calendars.
AI Systems
Tools that have shipped, broken, and been fixed in front of real clients.
How we look at it
Most rollouts fail in the gaps between People, Process and Technology because three different people own them and nobody owns the joins.
We own the joins.
01
People
02
Process
03
Technology
How it runs
Two weeks. We talk to the people doing the work. Watch a few workflows. Open the subscriptions list. Write you an honest read on where the leaks are.
Blueprints for the workflows that actually move the number. A short list of what the team needs to learn, in the order they will need it.
We sit with the team and build it. Prompts, small internal apps, SOPs. Nothing gets handed over cold.
Thirty days of office hours, async review on anything new the team ships, and a written-up version of the system they are now running on.
The other bit
With a bit of scaffolding, the person who needs the tool can ship it in an afternoon. We teach the team to do that properly. What to build, what not to touch, and where the guardrails sit.
The queue for engineering gets shorter. The work gets done.
How to brief the model
Like you'd brief a junior. Not like you'd Google something.
What to ship, what to escalate
Including how to keep client data out of the wrong tools.
How to read the output
Spotting the bits AI made up, and writing the correction.
What to leave alone
The decisions that have to stay human, and how to push back when a tool reaches for them.
What changes
People
Capability
Everyone on the team can use AI on their own work, brief it, and check it. No two-tier team.
Process
Cycle time
The workflows that matter run faster, without quality slipping.
Technology
Spend and risk
Smaller kit. Owners named. Governance written down, not folkloric.
Before you ask
Start a conversation
We read every one of these ourselves. You'll hear back inside two working days with a straight answer on whether we're the right call.