Set your dev team up to enjoy AI tools and multiply the output
Service for the CTOs and CIOs with passion and pressure to get their development teams boosted with AI. We know the resistance and how most of the teams don’t want to change. First we make sure the best practices and tools are available for the whole team. Then we work with the team and assist the first movers to utilize and further develop the project and repo specific tools for them to enjoy. Pretty soon they all discover and become inspired by the creative power that AI holds. The 90/9/1 process is industry tested.
€5,000 + VATFixed price, for up to 15 developers across up to 3 teams
- Face to face in the Nordics and Baltics (travel at cost)
- Delivered in English or Finnish
- Scales to larger organizations
Booking starts with a 30-minute call, not an invoice. If it isn’t a fit, we say so.
You bought the tools. Your team never changed how it works.
If you’re honest, day to day it probably looks like this:
- You’re paying for Copilot, Cursor, or Claude seats, and you can’t point to anything that got faster.
- One or two developers are genuinely flying with AI. The rest of the team works much the way it always has.
- The people who cracked it keep the good moves in their heads, and nobody else catches up.
- Every developer is working it out alone, reinventing the same prompts and hitting the same walls.
- The AI works on generic code, not yours, because nothing is connected to your systems or conventions.
- You can’t tell whether any of it is helping, because nothing is measured.
- There’s a new tool every month, you’re always a step behind, and nothing compounds.
- Leadership keeps asking what you’re getting from AI, and you still don’t have a real answer.
Meanwhile the roadmap is due, and the pressure to show something for AI is landing on you.
It isn’t you, and it isn’t your team. Adoption splits the same way everywhere.
As Angie Jones, VP of the Agentic AI Foundation, put it on the Practical AI podcast: in any group, 1% create, 9% tinker, and 90% consume. She used exactly that split to drive AI adoption across 3,500 developers, not by training everyone at once, but by turning the 1% of creators into champions the rest of the team could follow.
That split lives inside your team, not just across the market. Most of your developers are consuming or tinkering. One or two, your 1%, have worked out something that genuinely helps. But it stays trapped in their heads, because there is no system to spread it and no room for their voice to carry. So the gains never reach the team, and never reach you.Source: Practical AI, Aug 2026
We don’t train fifteen people one by one. We set up your champions, and it spreads.
The answer is not more tools, and it is not more training days. It is the move the 1% make: take what actually works, bake it into your systems, and let it reach everyone through the people who own the code.
So we work with your forerunners, turning your 1% of creators into champions the rest of the team can follow. If you already have that developer, the one who keeps finding a better way but can’t get it to stick, we amplify them and give their approach the reach it never had. If you don’t have a 1% yet, we act as it, and build your first champions with you.
Together with your champions we cut through the noise and find what genuinely works on your kind of code, not on a demo. Then we bake it into the systems themselves, the agent instructions, skills, context, and pipelines in your repositories, so any developer’s agent works the way your team wants. Your champions carry it to their teams, and the whole team benefits without everyone having to learn it from scratch.
- 1 yr
- of delivering agents
- 10+
- agents in production
- 100B
- tokens of code produced
We ship with AI every day. Now your team can too. We are that 1%: we’ve built this way of working for ourselves and for clients, so we’re not theorizing about it. We live there, and in the engagement we build it with your team, shaped around how you already work.
Who does the workYou work directly with Jonas Pomoell: a full-stack software engineer and business consultant who has been shipping software since 2008, for companies large and small. No junior bench, no handoffs, the person who scopes the engagement is the person who runs it. More about Jonas
Whatever your team builds, we have the tooling for it.
What works with AI depends on the repo and the use case: a website is not a mobile app, a backend service is not a data pipeline. So we bring the skills, plugins, processes, and MCPs, tailored to each one and set up where your team already works. Your use cases are covered.
- Websites and web apps
- Servers and APIs
- Mobile apps (iOS and Android)
- Data and infrastructure
Not on the list? We tailor the same skills and tooling to your kind of code all the same.
One engagement. Your team working in the AI era.
A focused, fixed-price engagement. We study your team, run the workshop, deploy the tooling into your systems, and set up your champions. You leave with a working, repeatable way of working, not a slide deck and a list of recommendations.
Measure it, don’t guess. If you already track how fast your team ships, we work against that baseline. If you don’t, we help you define a benchmark worth watching, so you can tell whether it is working over the weeks that follow, in numbers rather than a gut feeling.
Four moves, then your team runs with it.
- 01
Study
We interview your leads and developers and watch how the team actually works: where the time goes, what AI tooling is already in play, which repositories and workflows matter most, and the stack and security constraints everything has to fit. Not a questionnaire, real sessions with the people who write the code.
- 02
Workshop
A hands-on session with your champions, and anyone keen to become one. We cut through the noise together (there is a lot of it) and try things on your real repositories, so what we keep is proven on your code, not on someone else’s demo.
- 03
Deploy the tooling
We set up and make available to the team the pieces that turn AI from a toy into a tool: coding agents, isolated environments to run them safely, the plugins, skills, and MCP connections into your own systems and data, and the pipelines that tie it together. Baked into your repositories, so whoever points an agent at that code, it works the way your team wants.
- 04
Spread and measure
Your champions carry it into their teams the way that fits you: a brown-bag session, a shared doc, some pairing. We set up an iteration loop and something to measure against, so the gains keep compounding after we leave. From there your team explores and finds the value on its own.
We build on what you already run, not a platform we make you switch to.
- AI assistants
- Claude Code, GitHub Copilot, Cursor, Codex, Windsurf
- Version control
- GitHub, GitLab, Bitbucket
- Issue tracking
- Jira, Linear, Asana
- Testing and CI
- Playwright, Cypress, Jest, GitHub Actions
- Containers and cloud
- Docker, Kubernetes, AWS, GCP, Azure, isolated cloud runners
- Security
- secrets management, least-privilege access, audit logging, sandboxed execution
- Context
- MCP servers into your own systems and data
Whatever you run, we consolidate to what works, connect the rest, and run the coding agents safely inside your own environment. No rip and replace.
You finish owning the capability, not renting it.
- Champions equipped to lead your team’s AI-assisted development, and a shared way of working the whole team owns.
- The tooling deployed and available to everyone: agents, isolated environments, plugins, skills, MCP context, and pipelines, baked into your repositories.
- AI connected to your real systems and conventions, so it works on your codebase and not the open internet.
- An iteration loop and a benchmark worth watching, so the value compounds after we leave.
- A straight answer for leadership on where AI is paying off, backed by something you can measure.
Built for the leader whose team has AI tools but hasn’t changed how it works.
For engineering leaders (a CTO, a CIO, a VP of Engineering, a team lead) who are on the hook for getting real value from AI and aren’t there yet. The price is set for up to 15 developers across up to 3 teams: big enough that a shared way of working compounds, small enough that we reach every team. Smaller works too, it’s simply a lighter lift. If your team has barely touched AI, this is the right first step. If you already run a mature, measured AI-development practice, you don’t need us.
Bigger than that? A larger organization is a scoped rollout, not a single engagement, and it scales cleanly across more teams. We work in English or Finnish, and we come to you: face to face across the Nordics and the Baltics, with travel billed at cost. Further afield, we work remotely.
The questions engineering leaders actually ask.
- We already have Copilot and Cursor. Isn’t that this?
- Having the tools open is the 9%. It is not the same as a shared, measured way of working that reaches the whole team. That gap, scattered individual use versus a real team capability, is exactly what we close, and it is where the gains have been hiding.
- Do you write our code for us?
- No. This is consulting and enablement, not a body shop. We do not parachute in and write your code, because that capability would walk out the door when we do. We get your team, and its champions, working with AI so the capability stays with you.
- How much of my team’s time does this take?
- Most of the load sits with your champions and with us, not with the whole team. Expect a slice of your forerunners’ time for the study and the workshop, and short sessions for everyone else. Nobody has to down tools for a solid week.
- Is a short engagement really enough?
- Enough to make the change possible, not to do it for you. We remove what kept the team stuck and set up the champions and the shared system. The change itself happens inside your team over the weeks that follow, as your people run with it.
- Which tools do you use?
- Whatever fits your stack. We are tool-agnostic and consolidate to what actually works for your team and your constraints, not to a platform we happen to resell. On AI specifically we are deliberate: applied where it genuinely helps, not by default.
- Does it work with our setup and security rules?
- Yes. The study is built around exactly that: your stack, your systems, your constraints. Everything we deploy is shaped to fit them, the agents run inside your own environment, and MCP connections respect your access rules.
- Can you come on-site, and in which language?
- We work in English or Finnish. We come to you face to face across the Nordics and the Baltics, with travel billed at cost. Further afield, we work remotely.
- We’re bigger than 15 developers.
- Then it is a scoped rollout rather than a single engagement. The approach scales cleanly across more teams; the fixed price here is set for up to 15 developers in up to 3 teams. Tell us the shape and we will size it.
Give your whole team what your best developer already has.
€5,000 + VATFixed price, for up to 15 developers across up to 3 teams.