Services · Rollout Workshop, Team Launch, Care
Roll out agentic coding with the people and processes you already have.
No chaos, no disruption. Start with a one-day workshop, launch one workflow in your existing tools, then keep it current with monthly Care. We leave working software behind, not slides.
Your experts shape every agent's work.
Subject-matter experts become decision influencers: what they know goes into their own Brain Kitty, a knowledge graph of the standards, runbooks and decisions they own, which your agents read.
Your decision makers decide fast, fully informed.
They get the context to make AI-guided decisions quickly, without guessing what was already decided.
Who it's for.
Who books it
CTOs, VPs of Engineering and heads of engineering whose teams already use AI coding tools, but without a shared way of working around them.
Who is in the room
Product, engineering leads, reviewers, and the subject-matter experts whose knowledge your agents need. Up to 20 people. The day works best when the people who decide and the people who build are both there.
Three steps. Each one leaves something working.
Rollout Workshop
One remote day
Up to 20 people, remote, in your timezone.
- Your team's first Brain Kitty content
- Review gates for agent work
- A written rollout plan
Team Launch and Care are scoped with you on a fit call.
The Rollout Workshop
What the workshop day is like.
Up to 20 people, one remote day, finishing with a real item from your own backlog.
- One working day in your team's timezone: Europe, North America or APAC.
- Working blocks with real breaks, not eight hours of unbroken video call.
- Everyone on their own machine, with the AI coding tools they already use.
- We work in your repository, on your backlog. No toy examples.
- Team split across two timezones? We'll say on the fit call whether one day or two half-days works better.
The day, in outline
Morning
- Reviewing intent in plain language before anything is built
- Capturing your team's standards and decisions in Brain Kitty
- Planning work packages across parallel agents
Afternoon
- Working one real item from your backlog end to end, together
- Fitting the workflow to Jira or Linear without abandoning either
- Where the tooling stops, and what still needs a person
We adjust the order to where your team is and agree it with you before the day.
From a first workshop
“A ticket our own engineers had sized at ten days went out the door in about five hours — during the workshop, on our first time using it.”
Why it collapsed
Most of the ten-working-day estimate was agreement, not typing: three sprints of back-and-forth about what the ticket meant, who owned which decision, and what “done” included.
Putting the decision-makers in one room and writing the spec together decided all of that in a single pass. The agents then had an unambiguous brief, and the reviewer knew what to check.
One team's result on one ticket, not a benchmark.
For developers
Make the case to your manager.
Want this for your team? Here is the short version to forward.
- The problem: Our engineers use AI coding tools without a shared way of working, and what our senior people know and have decided rarely reaches the agents.
- What the day produces: One remote day for up to 20 people, in our timezone, on our own repository and backlog. We leave with our first Brain Kitty content, review gates for agent work and a written rollout plan.
- Cost: $3,450 · €2,950 · £2,550 for the day. The Spec Kitty CLI stays free and MIT-licensed.
- Proof: One team shipped a ticket its engineers had sized at ten working days in about five hours, during their first workshop. One team's result on one ticket, not a benchmark.
- Next step: A free 30-minute fit call to check whether it fits our team.
Ask about a package
Tell us which step fits. We reply within one business day.
