What every leader walks away with
Four things you can verify. If one is missing, the program is not finished.
-
They can explain what agents change in their area, and why waiting has a cost. Which tasks move to agents, which roles change, and what a manager manages when part of the team is agents. They can judge an AI proposal on its merits and explain the call to their own boss and peers.
-
Their own leader work runs on a second brain in Claude Cowork or ChatGPT Work. What they know about their team, the decisions and the reasons behind them, the promises they made, and the standard a result has to meet, kept in one folder the AI reads before every job. One-to-one prep, decision memos and the weekly team digest run on it, some on a schedule.
-
A ranked roadmap of what their team hands to AI. Every team workflow mapped, scored on AI fit and readiness, and given a type: assistant, automation or agent. Each of the top three has an owner and a risk level the company can sign off on.
-
One team agent specced, an owner named, and the team brought along. A one-page spec a team member can build from, a weekly routine where the team reviews what the agents did, and a plan for the conversation with the team about what changes.
Four weeks: what agents change, the leader's own work, the team's roadmap, and leading the team through it.
01
Understand the new AI workforce, and why it pays to act now.
- Understand what a team looks like when part of the work is done by agents.
- Build the AI and agentic fluency to judge proposals and lead with credibility.
- See how fast the work is changing, and what it costs a team to wait.
Understand the new AI workforce, and why it pays to act now.
- Understand what a team looks like when part of the work is done by agents.
- Build the AI and agentic fluency to judge proposals and lead with credibility.
- See how fast the work is changing, and what it costs a team to wait.
I can say what agents will change in my area, and I know which AI proposals deserve a yes.
They write down the three jobs in their area that take their team the most time each week. If an AI proposal or a vendor pitch is on their desk, they bring it.
One of the jobs they brought, done live by an agent in Claude Cowork or ChatGPT Work, in front of them. Then examples of what comparable teams already hand to agents, and what changes for the people around that work: which tasks go, which roles grow, what the manager now checks. Then the questions a leader asks of any AI proposal, applied to the proposals in the room.
They write a one-page brief for their own boss or peers: what agents will change in their area, what they want to try first, and what they need approved. They run one real proposal through the five questions.
Some of the work moves to agents. People move to setting that work up, checking it and handling the exceptions. The manager answers for both.
What does it do without a person? How was it tested? What data does it read? Who approves what it produces? What does it cost to run each month?
Every month a team waits is another month of the same hours spent on work an agent could already prepare. The live demo shows this on a job from their own area.
Deliverable A one-page brief for their boss or peers on what agents change in their area, and one real AI proposal assessed with the five questions.
02
Use AI for the work only a leader does, on a second brain in Claude Cowork or ChatGPT Work.
- Learn the best leader use cases: preparing decisions, one-to-ones, reviews and executive updates.
- Keep what they know about their area in a second brain the AI reads before every job.
- Put the recurring leader jobs on a schedule.
Use AI for the work only a leader does, on a second brain in Claude Cowork or ChatGPT Work.
- Learn the best leader use cases: preparing decisions, one-to-ones, reviews and executive updates.
- Keep what they know about their area in a second brain the AI reads before every job.
- Put the recurring leader jobs on a schedule.
When I prepare a decision or a difficult conversation, the AI already knows my team, my priorities and what I promised.
They list what they re-explain to the AI every time: the team, the priorities, the decisions and the reasons behind them, the promises made. Notes from their last three one-to-ones come too.
Six or seven documents about their area, written in the room into one folder in Claude Cowork or ChatGPT Work. Then three leader jobs run on it: preparing a one-to-one, drafting a decision memo, and turning the team's reports into a weekly digest. The same questions go to a blank chat for comparison. The digest and the one-to-one prep go on a schedule. Last, what goes into the second brain and what never does.
Before a real meeting, they put a real question to the second brain. The weekly digest arrives once on its own. They add their first lesson to the second brain.
The team, the priorities, the decisions, the quality bar, worked examples, and the vocabulary of the area. The rest gets added through use.
Decision memos, one-to-one prep, performance review drafts, board and executive notes, and reading what the team sends up. The work only the leader does.
The scheduled jobs log what they did into the second brain. The leader adds one lesson a week. That keeps it current.
This is each leader's own second brain, holding their data under their rules. A shared one for the whole team needs its own permissions and is a separate project.
Deliverable A second brain with six or seven real documents, three leader jobs run on it, the weekly digest and one-to-one prep on a schedule, and one real question answered before a real meeting.
03
Build the team's agentic roadmap: which work goes to AI first, and under which rules.
- Understand when a workflow needs an assistant, an automation or an agent.
- Map the team's workflows to find the strongest candidates for AI.
- Prioritize them by AI fit and readiness, with risk and governance decided up front.
Build the team's agentic roadmap: which work goes to AI first, and under which rules.
- Understand when a workflow needs an assistant, an automation or an agent.
- Map the team's workflows to find the strongest candidates for AI.
- Prioritize them by AI fit and readiness, with risk and governance decided up front.
I know what my team hands to AI first, second and third, and what each one needs approved.
They list the team's recurring workflows: who does each one, how often, how many hours, and where the inputs come from. They ask two people on the team what they would hand off first. If IT or security has a policy on AI tools and data, they bring it.
First, assistant, automation or agent, decided per workflow: an automation when the steps never change, an agent when the work needs judgment along the way, an assistant when a person stays in the loop at every step. Then the map: every workflow scored on AI fit and on readiness. Then risk: what data each one reads, what can go wrong, what cannot be undone, and what has to go to IT or security. The top three get an owner.
They finish the map for the whole team and share the top three with the people who would own them.
Is the work mostly reading, writing, sorting or comparing? Is a wrong answer cheap to catch before it does damage?
Are the inputs in one place? Can "done" be said in a sentence? Does one person own it? A workflow high on fit but low on readiness gets its inputs sorted out first.
Low: internal drafts a person reads. Medium: customer or personal data. High: it sends, pays or decides something that cannot be undone. The level decides who signs off.
Anything that needs a connector or new data access is planned against what the company has already enabled. We do not promise access that security has not approved.
Deliverable A ranked map of the team's workflows and a roadmap for the top three, each with its type, its owner and its risk level.
04
Spec the first team agent, and lead the team through the change.
- Spec an agentic team workflow that someone on the team can build.
- Embed AI accountability into the team's weekly routine.
- Overcome resistance: talk with the team about what changes and what does not.
Spec the first team agent, and lead the team through the change.
- Spec an agentic team workflow that someone on the team can build.
- Embed AI accountability into the team's weekly routine.
- Overcome resistance: talk with the team about what changes and what does not.
The first team agent has a spec and an owner, and my team reviews it every week like any other work.
They bring five real examples of the top workflow with the right result already worked out, and the objection about AI they hear most often from their team.
The spec, written live with five steps: decide whether the workflow needs an agent or fixed rules, test against the five examples, design the steps, equip it with instructions and tools, and protect it with the approvals set in week 3. Then the team routine: a standing item in the weekly meeting, one question for one-to-ones, and what to track. Last, the conversation with the team, practiced on the objections they brought.
The spec goes to its owner. The first weekly review takes place in the team meeting. They hand in the spec, the routine, and a short note on how the conversation with the team went.
The leader decides what the agent does, how it is tested and who approves. A named person on the team builds it in Claude Cowork or ChatGPT Work.
In the team meeting: which agents ran, what came out, who checked it, what to change. Each agent has one owner. Tracked every week: hours given back, errors caught, rework.
Say which tasks change, which roles stay, and what support people get. Answer objections with the team's own results from the first weeks.
The five steps come from the published guides by Anthropic (Building effective agents) and OpenAI (A practical guide to building agents), reduced to the decisions a leader makes without writing code.
Deliverable A one-page spec with a named owner, the weekly team review on the calendar, and a plan for the team conversation, tried at least once.
How the program runs
Live · 1 h 15 min
Enough to act on, then a demo on a real job from their area. The rest is work on their own area and their own team. The last 10 to 15 minutes are questions.
Async work · 1 h 15 min
Before the first session: the pre-assessment and the three jobs that take their team the most time. Between sessions: the second brain, the team's workflow list, and five worked examples for the spec. Real work from their own area, not exercises.
Before and after, on real leader work
No generic case studies. Every leader arrives with one of these, or with their own.
Their own work
Two hours of drafting the night before→15 minutes of edits on a draft that already sounds like them
Pulling numbers and past decisions out of five places→a draft with the context and the open questions, ready to edit
Rereading old notes and chats before each meeting→one page with what was promised, what moved and what to raise
Their team
Twelve reviews written from scratch every six months→twelve evidence-backed drafts, ready to review
Chasing everyone's progress through chat threads→the digest ready before the meeting, every week
14 files in 14 different formats→one comparable page the leader reads in five minutes
The team's roadmap
Three days pulling spreadsheets out of five departments→mapped, scored, and specced as the team's first agent
Everyone rules on them by their own judgment→a spec with the same criteria for everyone and a named approver
Approved or blocked on instinct→five questions answered on one page before anyone says yes
A job that is not on this list still fits. The mechanics are the same for any work a leader or their team repeats.
Who this is for
For
Directors, heads of area and managers with a team, who already use AI at a basic level in their own work and are comfortable in Claude Cowork or ChatGPT Work. Now they have to decide what their team hands to AI, judge the proposals that land on their desk, and bring their people along.
Not for
Anyone who has never used AI, who is better served by the AI Literacy program. People who want to build agents themselves, which is the AI Builder Bootcamp. And anyone who wants a talk without opening the tool.
How we run it
Enough to act on, then act. One short idea, one real demo, and the rest of the time on their own area and team.
Assess
A questionnaire for every leader before we start. Not a test. We ask how big their team is, how they use AI today, and which AI proposals are on their desk. The answers set the pre-work and calibrate the room.
Coach
Every session opens with one short idea and a demo on a real job. The rest is work on their own area and team, with us in the room.
Run
Between sessions the second brain runs on real leader work, and the roadmap gets built with the real team. Each week leaves behind something they can show.
From using Claude Cowork or ChatGPT Work yourself to leading a team that works with agents
Today: each person tries AI on their own
- Everyone on the team uses AI their own way, and nobody has listed which work should go first.
- AI proposals get approved or blocked on instinct.
- Nobody reviews what the agents produce, or knows who answers for it.
- The team hears about AI through rumors, not from their manager.
After: a team roadmap with owners, rules and a weekly review
- A ranked map of the team's work, and the first agent specced with an owner.
- Risk levels and approvals decided before anything runs.
- A weekly review where the team checks what the agents did.
- A second brain the leader uses before every decision and difficult conversation.
It does not stop at the last session
Certificate
A shareable certificate once the four weeks are done and the spec is with its owner.
Opportunity report
A report built from the ten team roadmaps: which workflows came up across teams, which are ready now, and which need IT or security first. A set of next steps for the company, not a satisfaction survey.
Platform access
Exercises, materials, second brain templates and team roadmaps stay on the AdapttoAI platform after the program ends.