How it works
What seven weeks with us looks like.
This page is the plain-language version of our statement of work. It is the structure we have run inside several Sacramento companies this year, refined each time. If you would rather read the actual contract before we talk, ask and we will send it.
Weeks 1 and 2 · Preparation
Two weeks of preparation set the agenda.
Before the first workshop, we work with leadership and IT so day one is not a setup day. We ask for two things. Leadership fills out an AI policy questionnaire: which tools are approved, what data stays off-limits, who reviews what. Each participating team member fills out a survey listing their recurring tasks, scored for how tedious they are, how long they take, and how much they matter to the company. We also collect sample documents, the real ones, and confirm everyone has a paid seat on the platform you chose.
What comes out of it
- A survey dataset for all participants, in a format you can use
- A findings report: the highest-drudgery, highest-impact tasks, who is ready for what, and the order we recommend
- A draft AI use policy for your company, written with leadership and ready for your attorney
- Accounts, seats, and connectors set up by our IT team so everyone is ready on day one
Weeks 1 and 2 · The findings report
Which workflows should become AI skills?
This is what the findings report does with your survey. Every recurring task in an office sits somewhere on this map. Across is how many hours a week it eats, on a square-root scale so the small tasks stay legible. Up is how much it matters plus how tedious it is, each scored 1 to 3. Dividers sit at 8 hours a week and at a combined score of 4.5. Tap any circle, or plot one of your own.
Plot one of yours
Pick a task someone on your team does every week and see where it lands.
Sample workflows from the kind of office team we usually train. On a real engagement the scores come from the people who do the work, not from us.
Weeks 3 to 7 · The five workshops
Two hours, one morning a week, at your office.
Workshops run on the same day and time each week, usually a Monday morning, for about two hours. They are the part of the program the whole team attends together; the forty hours of skill building below happen at people's desks in between. The outline is what we have delivered; topics shift when a vendor ships something useful or when your team needs more time on one thing.
Hover or tap any bar to see what happens that week.
-
01
Week 3: Get hands-on
Full team. Prompting fundamentals, research, and a first look at the tools' document and file features. Everyone works through a real task before leaving.
-
02
Week 4: The AI clinic
Bring real work: RFIs, change orders, estimates, reports, emails. One-on-one and small-group coaching on each person's highest-priority tasks.
-
03
Week 5: Workflows and capstone lock
Building reusable projects and skills. Each team member commits in writing to the one workflow they will build.
-
04
Week 6: Capstone build
Entirely dedicated to finishing capstones, with coaching from us in the room. Office Hours between workshops for one-on-one help.
-
05
Week 7: Presentations
Each person presents their capstone using a template we provide. Peer vote, leadership pick, small prizes, and handoff of all materials.
Between workshops · Build at the desk
Forty hours with our workflow engineers, in two buckets.
The workshops teach. The hours between them are where the work actually gets built: our workflow engineers sit with people one-on-one or by department, at their own computer, on their own work. We keep the two kinds of hours separate on purpose: if they were one pool, capstone deadlines would eat all of it and the shared tools would never get made.
Office Hours (20 hours)
Reactive. A team member brings what they are building and gets help with it. One-on-one or small group, scheduled around your team, in person or on a call.
Collaborative Skill Building (20 hours)
For the shared assets no individual capstone covers: writing down the procedure that lives in someone's head, building a tool several people will reuse, setting up a workflow that spans roles. We build with your people, not for them, so they can maintain it.
What is guaranteed
Three written outputs. Everything else is built with you.
We want to be precise here, because it is the part most vendors blur. What we commit to under the contract is the workshop time, the forty hours of skill building at the desk, and the policy and IT setup sessions, delivered by us, plus the three written outputs from the surveys. Everything your team builds during the program, the prompts, projects, skills, workflows, and capstones, is a collaborative result. It depends on your team's participation, their knowledge of the work, and access to your systems. We do not promise that any particular automation gets finished. We do promise that the people building it are in the room.
- Survey dataset for every participant
- Findings report on where AI is most likely to pay off for your team
- Draft AI use policy for your company
- Twelve months of email support after the program ends
- A one-hour check-in about ten weeks after we finish, to review what stuck
What we need from you
A room, a point of contact, and your real documents.
- One person to coordinate scheduling and materials
- The two surveys completed before Session 1
- Sample documents: forms, procedures, estimates, reports, whatever your team actually produces
- Paid seats on Claude or ChatGPT for everyone participating
- The team in the room for all five workshops
- A meeting space at your office
Pricing
$15,000 program fee, plus $500 per participant.
The program fee covers everything that costs the same whether nine people or nineteen take part: the two weeks of preparation with leadership and IT, the findings report and policy draft, the five workshops, the forty hours of skill building with our workflow engineers, and twelve months of email support. Each participant adds $500 for their survey, their clinic seat, and their capstone coaching. You choose who participates. Most companies send the office team: estimating, project management, accounting, admin. Field crews do not need to attend.
| Participating team | Program total |
|---|---|
| 8 participants | $19,000 |
| 10 participants | $20,000 |
| 15 participants | $22,500 |
| 20 participants | $25,000 |
| Larger teams | We run it in two groups so the clinic stays hands-on. Ask us. |
Common questions
Do we need to pick Claude or ChatGPT first?
No. Both work for the program. We help you decide before Session 1 based on what your company already runs, and we do not take referral fees from either vendor.
What if some people on the team are skeptical?
Good. The surveys tell us who is, and we plan for it. A skeptic with one small, tedious task often gets the best first win in the building.
Is my team too non-technical?
No. The teams we have trained were estimators, office coordinators, accountants, and lobbyists, not engineers. If someone can write an email, they can do this program.
What happens to our data?
Your leadership sets the rules in the policy questionnaire, and we teach to them. Under the business plans of both vendors, your content is not used to train their models by default, and we confirm those settings on your plan. Anything you mark off-limits stays off-limits.
Who counts as a participant?
Anyone who attends the workshops and builds a capstone. Most companies send the office team and leave field crews out. We help you decide before we quote.
Can the program be shorter?
We have run a shorter version. The price works the same way: a program fee plus participants.
What happens after week five?
Twelve months of email support, and a one-hour check-in about ten weeks out to see what stuck. If you want ongoing build help after that, we can scope it separately. It is not part of this price.
Who actually shows up?
Tom runs every workshop and is your point of contact from the first email to the check-in. Allen leads the workflow engineers and the IT setup. You meet everyone who will be in your office before you sign.
Want the contract?
Email us and we will send the statement of work. It reads like this page.