SMALL BUSINESS & MID-MARKET
The teams with no internal trainer
For operations, finance, HR, and customer teams that have access to AI but need practical guidance and role-based examples.
TRAINING
Help your team get more from the AI tools it already uses.
Training is built around the work your people do, not a made-up example. Sessions are available for Microsoft 365 Copilot or Claude and can be shaped to your team’s roles, tools, and starting point.
USE WHAT YOU HAVE
Use the tools you already have.
Many organizations already have access to an AI assistant. Training can help staff use it more effectively before you consider adding another tool.
STILL INDEPENDENT
No vendor commissions, no reseller badge.
I take no commissions, referral fees, or partner incentives from Microsoft or Anthropic. If a tool is not right for a workflow, I’ll say so. The skills your team builds should transfer to other assistants too.
SERVICE 02 · TRAINING
AI training, sized to your team
At any length, the training starts with tasks your team brings to the room. Participants practice themselves and leave with notes and examples they can use again.
Training shaped around real work
- Start from the work: use real tasks from real roles.
- Practice hands-on: participants work directly in the tools available to them.
- Use clear guardrails: agree what information can be entered and how to check the output.
- Measure progress: compare task time and quality before and after where practical.
- Keep the materials: receive session notes, role-based examples, and documentation for future staff.
A six-session program
The full program is six sessions, about two hours each, roughly 27 hours in total. The final session focuses on completing and reviewing a workflow from each participant’s own job.
Microsoft 365 Copilot track:
- Session 1. Chat, working with your own files, Pages, Notebooks, and choosing a model
- Session 2. Word and Outlook
- Session 3. Excel
- Session 4. Teams, PowerPoint, and OneNote
- Session 5. Agents
- Session 6. Complete, document, and review a workflow
Claude track:
- Session 1. Chat, Projects, working with your own files, and choosing a model
- Session 2. Drafting and documents
- Session 3. Data, spreadsheets, and analysis
- Session 4. Connectors and the tools your organization already uses
- Session 5. Agents and automation, including Claude Code for technical teams
- Session 6. Complete, document, and review a workflow
The full program includes a documented workflow for each participant, a review against a shared rubric, and a certificate of completion. It also includes one free refresher session within twelve months of the final session.
Choose the right length
- Shorter. A briefing, single session, half-day, full-day, or shorter multi-session series can introduce useful habits or focus on a specific need.
- Longer. Run the program by department, add sessions for new employees, or schedule follow-up sessions as tools change.
- Tailored. Build the sessions around workflows your leaders have identified, rather than using the same lesson plan for every team.
A shorter format will not produce every deliverable included in the full program. Ninety minutes can change habits, but it will not produce a documented, reviewed workflow for every person in the room. A short call is the best way to decide which length fits.
WHERE YOUR PEOPLE ARE
Different starting points
Teams often include people with a range of experience. Training can meet people where they are and help them take a useful next step.
Curious
Has tried an assistant but has not made it part of regular work.
Deliberate
Uses it for a few specific tasks.
Daily
Uses it across several tasks or apps.
Delegating
Hands over larger pieces of work and reviews the result carefully.
Sharing
Has built and documented a workflow other people can use.
A full program is built to move people toward level five. That is why the last session is about completing and reviewing a workflow, not adding more features.
THE HONEST PRINT
What gets taught is re-checked before it gets taught.
These assistants change every month, so a course written once and never rechecked is wrong within a quarter. I check this material against the vendor’s own release notes before every delivery. Microsoft’s for Copilot, Anthropic’s for Claude. Every product claim carries a date and a source. Anything I cannot source does not get taught as fact, and I will not quote you a productivity number I cannot show you the origin of.
FIT
Who the training serves
Every session is taught by me, not a rotating bench of contractors. The background behind that sentence is on the About page.
SCHOOLS & FACULTIES
Admin load first, classroom second
For educators and administrators who want to use AI thoughtfully in planning, correspondence, reporting, and other administrative work.
PUBLIC SECTOR
Inside the guardrails you already have
For staff who need training that respects records rules, privacy requirements, and limits on the information they can enter into an assistant.
NORTHERN & INDIGENOUS ORGANIZATIONS
Capacity that stays after I leave
For teams that need training shaped around local governance, bandwidth, language, data control, and available capacity.
ONE REQUIREMENT
Participants need access to the assistant used in their sessions.
That means either Microsoft 365 Copilot or Claude. If you are still deciding what to buy or how many licences to purchase, start with an AI Readiness Assessment.
BEFORE YOU ASK
Questions that come up on every scoping call
Can you do this in less than six weeks?
Yes. Six sessions is the reference shape, not a fixed package. A single session works for orientation and one new habit each. A half day or full day can cover the apps your people live in. Two or three sessions cover a lot of ground when six weeks is more calendar than you have. What a shorter run costs you is the proof. A documented, measured, marked workflow for every single participant needs the full length. I would rather tell you that up front than sell you ninety minutes and call it a program.
What if we do not have Copilot licences yet?
Then this is not the right first step. Training assumes the licence is already on the desk, whether that is Copilot or Claude. If the open question is which assistant to buy, how many seats, or for which roles, that is the two-week AI Readiness Assessment. It answers the buying question independently, before you commit the money. The training conversation gets much easier afterwards. By then we already know which workflows are worth the time.
We use Claude, not Copilot. Does the program change?
The shape does not change; the apps do. The Claude track runs the same six-session arc on Claude's tools. Chat and Projects grounded on your own files, drafting in artifacts, data and analysis, connectors to your mail and calendar, then agents and automation. Claude Code comes in where a team is technical. Everyone still practices on their own work and still measures the task at both ends. A full run still finishes with a documented workflow marked against the rubric.
Do you train on tools other than Copilot and Claude?
Those two are the anchors because they are the licences organizations most often already hold and under-use. The method is the same whichever assistant you run. Sort what may and may not go in, ground it on your own material, practice on real tasks, measure both ends, and document what worked. I take no commissions or partner incentives from any vendor. So where the tool on your desks is the wrong answer for a specific workflow, you will hear that in the room.
Is this virtual or in person?
Both work. Live virtual sessions are the default. They work well because everyone is already in their own tenant with their own files open, which is where the practice has to happen anyway. In-person delivery is available. It is often the better choice for a first session, or for teams that are not comfortable learning on camera. Travel is quoted separately when it applies.
What do participants leave with?
At minimum, a prompt library shaped to the roles in the room and a habit they did not have before. On a full run, each person also leaves with a working, documented workflow from their own job. They get a run sheet another person could follow, a before-and-after timing on that task, and a certificate of completion.
How do you keep the content current?
Both assistants change monthly. I check the material against the vendor's own release notes before every delivery, Microsoft's for Copilot and Anthropic's for Claude. Every product claim carries a date and a source. Anything I cannot source does not get taught as fact. This matters more than it sounds: a course written once and never rechecked is wrong within a quarter. It is also why every full six-session run includes one free refresher session within twelve months. Your team gets caught up on what has changed since we finished.
LET’S TALK
Bring one real task to a 15-minute call.
We can discuss whether Copilot or Claude can help, what the training should cover, and what would count as a useful result.
Microsoft, Microsoft 365, and Copilot are trademarks of the Microsoft group of companies. Claude and Anthropic are trademarks of Anthropic, PBC. Travis Consulting is an independent practice and is not affiliated with, sponsored by, or endorsed by Microsoft or Anthropic. Product names are used only to identify the tools trained on.