AI Enablement Your Clients Can Actually Use
Usage training, use-case discovery, prompt design, and agent builds for the clients you already serve. You keep the relationship. We bring the course kits and the method.
Three Course Kits, One Method
Each kit ships as a facilitator guide and a participant workbook. Pick the level that fits the room. Every level teaches the same way of working.
Level 100 · 1 Day
Claude Essentials
Zero to productive in a day: workspace setup, outcome briefs, context and memory, a first connector, an automation showcase, and a retrofit express for the workflows people already run.
Attendees leave with
A participant workbook, a printable quick reference, and one written outcome brief for a task from their own week.
Level 200 · 3 Days
Claude Team Operator
Build the operator, build the system, build the machine. Day 2 is a full retrofit clinic on the team's real workflows. By the end of day 3 every team has a scheduled automation running.
Attendees leave with
A working scheduled automation (the capstone pulse), a Keep / Complement / Consolidate / Retire matrix for their workflow inventory, and facilitator-reviewed outcome briefs.
Level 300 · 4 Days
Claude Automation Architect
Everything in the Operator course plus a dual-track day: coding-agent depth for technical staff, desktop workflow mastery for operators. Finishes with plugin packaging, a governance one-pager, and train-the-trainer with a teach-back rubric.
Attendees leave with
Packaged, shareable skills, a one-page governance summary for leadership, and an in-house trainer who can run Level 100 without us.
Built on Claude. Delivered Where Your Client Already Works.
The kits use Claude as the reference platform because that is what we run our own business on. The method transfers. The exercises get rebuilt against the platform your client has already chosen.
Claude
Reference platform for every kit: chat, desktop agent, and coding agent, with the connector and scheduling features the capstone automations are built on.
OpenCode
Open-source coding agent for teams that want model choice and no vendor lock-in, including local and self-hosted models where the data cannot leave.
Gemini
For Google Workspace shops, where the mail, calendar, and documents the client wants to automate already live.
Microsoft Copilot
For Microsoft 365 shops, where Word, Excel, Outlook, and Teams are the daily tools and the client's tenant already sets the data boundaries.
The Method Underneath Every Course
What a client walks away with is a way of working, not a list of tricks. These three devices appear at every level, and they are the same ones we run our own business on.
The Maturity Ladder
Six rungs. Most teams are stuck on the first one. Each course moves the room at least two rungs up and shows them what the next one looks like.
-
1
Prompting
Asking well, once.
-
2
Context
Briefs, memory, and preferences that make every answer start from the right place.
-
3
Skills
Reusable, versioned instructions for the jobs the team repeats.
-
4
Connectors
Mail, calendar, tickets, and documents reachable from the same place.
-
5
Automation
Scheduled runs with approval gates on anything that sends, spends, or deletes.
-
6
Distribution
Packaging what works so the next team gets it without starting over.
Two Tracks in Every Room
Mixed technical and non-technical audiences are the norm for an MSP client, so every course runs both tracks side by side.
Greenfield
New workflows, built right the first time
Write the outcome brief, then build to it. Nothing gets automated until someone can say what done looks like.
Retrofit
Complement, don't replace
Inventory what runs today, set layering rules for where AI is allowed to sit, then sort every workflow into one of four bins.
The Outcome Brief
Every exercise starts with one. If the brief cannot be written in five minutes, the automation should not be built yet.
Outcome brief · template
- Outcome
- One sentence. What exists at the end that did not exist before.
- Audience
- Who reads or uses it, and what they already know.
- Inputs
- The data, documents, and systems the work is allowed to draw on.
- Done looks like
- The check a human runs before trusting the result.
- Guardrails
- What it must never do, and who approves anything that sends, spends, or deletes.
- Review cadence
- When someone re-reads the brief and decides whether the automation still earns its keep.
Train the People Before You Trust the Platform
Most AI rollouts fail at the keyboard, not in procurement. Staff get a login, try two prompts, get a mediocre answer, and quietly go back to the old way, or to a personal account nobody is governing. Usage training fixes the habits: how to brief, how to give context, when to trust an answer, and when to check it.
Delivered on site or remote, from a one-day essentials session to the four-day architect kit, to rooms that mix technical and non-technical staff. Your client's governance decisions become the first slide, not an afterthought.
Good for:
- Onboarding a client's staff after an AI readiness or data governance engagement
- Replacing shadow use of personal AI accounts with sanctioned, taught use
- Leadership briefings on what is realistic this year and what is not
Find the Hours Before Picking the Tool
"What should we do with AI" is the question every client asks and almost nobody can answer from a product demo. Discovery is a structured workshop with the people who actually do the work: inventory the workflows, score each one on frequency, time, error cost, and data sensitivity, and shortlist the handful worth a pilot.
The output is a prioritized use-case register and a 30-day pilot plan with an outcome brief per pilot. It is written so you can quote against it, and so the client can kill a platform spend that the numbers do not support.
Good for:
- Clients who want AI but have not named a single specific task for it
- Justifying, or declining, a platform purchase before the contract is signed
- Turning a governance policy into a short list of permitted, specific uses
Prompts That Outlive the Person Who Wrote Them
Every team has one person who is "good at AI." When they are on vacation, the quality drops. Prompt design turns that individual skill into shared infrastructure: a prompt library organized by role, templates tied to real workflows, and preferences and memory set up so results stay consistent from one staff member to the next.
Everything is versioned in a place the client controls, with maintenance notes, so the library keeps working after the platform updates and after we leave.
Good for:
- Consistent client-facing outputs: reports, summaries, status updates, replies
- Teams that re-prompt the same task from scratch every time
- Encoding the client's tone, terminology, and compliance language once instead of per person
Agents With Approval Gates, Not Autopilot
An agent is a scheduled job that can read the client's mail, calendar, tickets, and documents and act on them. That is useful and dangerous in equal measure, which is why every agent we build has a human approval step on anything that sends, spends, or deletes, logs what it did, and ships with a one-page governance summary the client's leadership can read.
We run our own business this way: our weekly marketing pulse and vendor brief are scheduled agents, and the course material teaches from what has broken in them, anonymized. Where data cannot leave the client's environment, we scope private or self-hosted model options rather than pretending the constraint is not there.
Good for:
- Recurring reports, briefs, and status summaries that someone currently assembles by hand
- Inbox and ticket triage with drafted replies that a human approves before anything sends
- Agents that must respect a client's data boundaries, including regulated and insurance-questionnaire environments
Built for MSPs, Not Around Them
You have already done the AI readiness, security, and data governance work with your clients. This is the delivery layer on top of it, and it stays in its lane.
You keep the client
We work through you, under your brand if you want it. We do not market to your clients, and the relationship, the renewals, and the credit are yours.
Your groundwork is the foundation
The governance policy, the sanctioned platform, and the data boundaries you set with the client become the guardrails every exercise and every agent is built inside.
Scoped so you can quote it
Per-course delivery, per-client discovery workshops, or a monthly enablement retainer.
Compliance-aware by default
We are a cybersecurity compliance firm first. Every engagement leaves behind artifacts a regulator, an auditor, or a cyber insurance underwriter can read: the brief, the guardrails, the approval log.
Frequently Asked Questions
Do you only train on Claude?
No. The course kits are built on Claude as the reference platform because that is what we run our own business on, but the method underneath them (outcome briefs, the maturity ladder, the retrofit matrix) is platform-agnostic. We deliver on Claude, OpenCode, Gemini, and Microsoft Copilot today, and rebuild the exercises against whatever platform your client has already standardized on.
Will you contact our clients directly?
No. We work through you. Engagements can be delivered under your brand, your client relationship stays yours, and we do not market to your clients. Your AI readiness, security, and data governance work is the foundation each engagement starts from.
Is building agents an IT job or a programming job?
Both, and the line is blurring. Every course runs a technical track and an operator track in the same room. Non-technical staff can build and own scheduled automations with approval gates; the Automation Architect kit adds a coding-agent track for technical staff who want to go deeper.
How do you handle client data during training and agent builds?
Every engagement starts from the governance baseline you have already set with the client. Exercises use the client's sanctioned platform and data boundaries, agents get human approval on any step that sends, spends, or deletes, and where data cannot leave the environment we scope private or self-hosted model options.
Related Services
Have a Client Asking About AI?
Tell us what they asked for. We will scope the training, the discovery workshop, or the agent build, and you deliver it as yours.
Schedule a Partner Call