01
Foundations
Understand AI, Claude, and the working modes. Demystify the technology and learn to interact with an LLM effectively.
AI Training plan
Master AI
as an operational advantage.
Three half-days to understand the full AI ecosystem — and walk away with everything you need to think, decide, and act on your own. In person, at your office. From theory to action.
Overview
Three sessions
to understand it all.
Each half-day builds on the previous. We start with AI fundamentals, move into team collaboration, and finish with advanced capabilities. By the end, your team understands the full ecosystem — and can decide knowingly how to go further.
02
The team brain
Obsidian as memory, Cowork as assistant, MCP to connect AI to your tools. Structure and connect your organization's intelligence.
03
The next level
Claude Code, the terminal, VS Code. Deploy your first agent and understand what becomes possible — and how to get there.
AI isn't a shortcut. It's a discipline. Well-designed, it frees your teams. Well-taught, it makes them autonomous.
Louis-Charles Cuierrier
Detailed program
What your team
will learn.
Each ~4-hour session combines theory, live demos and hands-on workshops in pairs. The goal: every participant leaves with a complete understanding of the ecosystem — and a functional agent at the end.
Important note
This plan is a proposal, not a rigid script. It adapts to your team's profile — a group already familiar with Claude doesn't need the same time as a team just discovering it. Modules added, removed, deepened or shortened based on your needs, sector and existing expertise. We calibrate together before the training.
Day 01
Foundations
The first session lays the groundwork. We demystify AI, understand how an LLM really works, and learn to work with Claude effectively. It's the foundation everything else builds on.
Understanding AI and LLMs
~45 min
What an LLM is (and isn't). How it works under the hood — no jargon. Why Claude stands out. The limits and strengths to understand for proper use.
LLM vs traditional AIHallucinationsContext windowTemperature
Contexts, memory and .md files
~60 min
How Claude handles conversation context. The memory system, projects, and why how you structure your exchanges changes the quality of results entirely. We also explore Markdown: why it's the universal format to communicate with AI, how a simple text file becomes your agents' persistent memory.
Conversation contextPersistent memoryClaude projectsMarkdown filesEffective prompting
The three working modes
~60 min
Chat, Cowork and Claude Code — three interfaces, three uses. We explore each in detail: when to use which, for what, and why understanding the difference matters.
Chat — direct conversationCowork — desktop agentClaude Code — terminalWhen to use what
Responsible AI — fundamentals
~30 min
AI without a framework is a risk. We explore 6 responsible AI principles: reliability, privacy, fairness, transparency, accountability, inclusivity. Concrete questionnaires your leadership can fill out to define the rules each AI agent will automatically respect.
6 Microsoft principlesHallucinations and reliabilitySensitive dataSTRICT BOUNDARIESGovernance for SMBs
Workshop — Same Claude, two realities
~45 min
In pairs, participants ask Claude 4 questions without context. Then they create a project, load a fictional company's .md file (provided), and ask the same questions again. The difference is immediate. They finish by editing the .md live to change Claude's behavior.
Pair exerciseWith vs without .md contextHallucination vs precisionEdit .md = edit behavior
What the team takes away
A solid understanding of AI, Claude, Markdown files as the foundation of AI memory, the different working modes, and responsible governance principles.
Homework before Day 02
Name the AI governance committee and fill out the principles
- Leadership identifies 2-3 people who will form the company's AI governance committee.
- The committee fills out the 6 responsible AI questionnaires (files provided). Each takes about 15 minutes.
- The answers become the company's STRICT BOUNDARIES, ready to integrate into the architecture on Day 02.
The 6 files are provided at the end of Day 01. What matters is that they're done before Day 02.
Day 02
The team
brain
brain
The second session gets into the concrete of collective intelligence. How to give your team a shared memory with Obsidian, and how to structure Cowork agents with defined roles and contexts.
Obsidian — the team LLM wiki
~75 min
Why Obsidian is the ideal tool to serve as the "brain" for your AI agents. Markdown file structure, internal links, and how to organize company knowledge so it's exploitable by AI.
Markdown and structureVaultsInternal linksKnowledge organizationTemplates
Cowork — the desktop agent in depth
~75 min
How Cowork turns Claude into an autonomous desktop assistant. Understanding roles, instructions, agent memory, and how a "Cowork team" lets each company function have its own specialized assistant.
Roles and instructionsAgent memoryCowork teamFlow automation
The Obsidian + Cowork architecture
~60 min
How the two work together. Obsidian as knowledge base, Cowork as action interface. The architecture that lets AI know your company and improve over time.
Agent-wiki architectureCumulative memoryShared knowledgeEvolution over time
Skills, Connectors and MCP
~60 min
How to extend your agents' capabilities. Skills teach a know-how. Connectors plug your agents into existing tools via MCP: Gmail, CRM, calendar, drive. Live demo of a Gmail and Calendar connection.
Custom SkillsNative ConnectorsMCP protocolGmail + Calendar demoSecurity and permissions
Workshop — Your agent connected to your tools
~45 min
In pairs, participants connect Gmail and Google Calendar to a Cowork agent: summarize emails, identify actions, check availability, draft a response. Everything stays in read or draft mode. The agent prepares, the human validates.
Pair exerciseGmail + Calendar connectorsCowork + MCP in actionAI prepares, you validate
What the team takes away
Complete understanding of how to structure a team's collective intelligence. Obsidian as memory, Cowork as action engine, MCP as bridge to your existing tools.
Homework before Day 03
Identify your agents and gather your sources
- Each department identifies 1-2 roles that would benefit most from an AI agent.
- For the priority role, gather 5-10 source documents: emails, templates, procedures, examples.
- Quick review of the STRICT BOUNDARIES filled out in the previous homework.
This reflection work directly prepares Day 03. The documents gathered will become the first sources ingested by your agents.
Day 03
The next
level
level
The third session starts with action: we deploy the agent you designed in your homework. Then we open the door to advanced automation with Claude Code, VS Code and the complete vision of the ecosystem.
Workshop — Deploy your first agent
~60 min
In pairs, participants deploy the agent they designed in their homework. We create the project, write instructions, load 2-3 source documents, and test with real cases. By the end of the hour, each pair has a functional agent.
Pair exerciseReal agent, not fictionalInstructions and contextSource ingestionFunctional result
Claude Code — AI in the terminal
~45 min
Introduction to Claude Code: how it works, what it enables, and how it turns a .md file into concrete action. The link between CLAUDE.md and operations Claude Code executes. No need to be a developer.
Terminal and command lineCLAUDE.md as instructionsFrom text file to executionConcrete use cases
VS Code, Git and the work environment
~30 min
VS Code as the environment to work with Claude Code. Git for versioning your files. Opening a folder, integrated terminal, viewing change history.
VS Code interfaceIntegrated terminalGit — basic versioningClaude Code in VS Code
Workshop — Hello Claude Code
~45 min
In pairs, participants open VS Code, launch Claude Code in a pre-configured folder, and run guided commands. They say "execute" and watch Claude Code read the CLAUDE.md and build a mini-project automatically.
Pair exerciseVS Code + terminalCLAUDE.md in actionAutomatic scaffoldFirst Git commit
What now? — the complete vision
~30 min
Synthesis of the whole ecosystem. How Chat, Cowork and Claude Code fit together. Presentation of the company vault architecture for those who want to go further.
Company vault architectureThink / Build / OperateAutonomy vs accompanimentNext steps
What the team takes away
A first functional agent working for your company. Understanding of Claude Code and VS Code. The complete vision of an evolving AI ecosystem.
The ecosystem
Eight concrete tools,
one single ecosystem.
Your team walks away with a complete understanding of each tool, its role, and how it fits into the whole.
Claude.ai
Main interface — chat, projects, memory.
Cowork
Desktop agent — flow automation.
Claude Code
Advanced automation — terminal and scripts.
Obsidian
LLM wiki — team memory, agent brain.
MCP
Connection to your existing tools.
Skills
Custom instructions for Claude.
VS Code
Development environment for agents.
Terminal
Total control — automation without limits.
Governance
Responsible AI,
built in by design.
The training includes a complete governance framework based on 6 recognized principles. Your team walks away with the tools to define the rules each AI agent will automatically respect.
Principle 1
Reliability & safety
AI works as expected. When it doesn't know, it says so. Zero fact fabrication, zero silent hallucinations. Sources are always cited.
Principle 2
Privacy & data
Sensitive data is identified and protected. Your agents know what they can see. Compliance with Law 25, PIPEDA.
Principle 3
Fairness
No discrimination, no hidden bias. A human always has the final word on decisions affecting people.
Principle 4
Transparency
The user knows when AI is involved, how it produced its answer, and where its sources come from.
Principle 5
Accountability
Each agent has a human owner. The AI prepares and recommends. The human decides and owns.
Principle 6
Inclusivity
AI serves everyone, regardless of language, technical level, or accessibility needs.
Recommendations
How it works concretely.
Each principle translates into a simple questionnaire your leadership fills out. The answers become STRICT BOUNDARIES that each AI agent inherits.
You fill them out once. Every agent inherits them. No per-agent configuration. No risk of forgetting a rule.
1
Fill out the 6 questionnaires (15 min each, guided by the AI advisor).
2
The answers become concrete rules (STRICT BOUNDARIES).
3
The rules are integrated into the company brain.
4
Each agent automatically inherits all the rules.
Format
How the training
unfolds.
Every detail is designed to maximize your team's learning and engagement.
i.
In person, at your office
The training is delivered directly at your offices. No travel for your team, no external logistics.
ii.
Groups of 8 to 12
Large enough to spark good discussions, small enough that every participant can ask questions.
iii.
Theory + live demos
Each concept comes with a real-time demonstration. No endless slides — we show tools in action.
iv.
Flexible sessions
The 3 half-days can be scheduled according to your reality — consecutive or spread out.
Prerequisites
What you need
before the training.
Practical workshops require each pair to have access to the tools. Here's what needs to be in place before Day 01.
Essential
Claude Team — 1 seat per pair
Each pair of participants shares a Claude Team account during workshops. For a group of 12, you need 6 seats. The Team plan gives access to everything: projects, Cowork, Claude Code, Skills and Connectors.
The seats remain after the training. It's an investment, not an expense.
6
seats for 12 people
$30
USD / seat / month
One laptop per pair
Mac or Windows. Internet required.
Day 01-02-03
Obsidian installed
Free. obsidian.md — 2-minute install.
Day 02
Gmail account accessible
For the MCP workshop.
Day 02
VS Code installed
Free. code.visualstudio.com.
Day 03
Git installed
Free. Often already on Mac/Linux.
Day 03
Claude Code installed
Requires Node.js. Install guide provided 1 week before.
Day 03
Included with trainingInstall guide PDF sent 7 days before. Pre-configured workshop environments. Email setup support if needed.
Who it's for
Who should
be in the room.
This training is for teams who want to understand AI to use it — not just hear about it.
Profile 1
Overloaded technical teams
Telecom resellers, integrators, manufacturer reps — teams whose experts spend too much time on administrative, repetitive tasks. AI can give them back time for what matters.
Profile 2
High-performance SMBs
Entrepreneurs, trades, business owners. Those who do everything themselves — quotes, invoices, follow-ups — and spend evenings catching up on admin. AI as virtual assistant.
Profile 3
Leaders & managers
Those who need to understand what AI can do before making investment decisions. Not buzzwords — real understanding.
Profile 4
Mixed teams
The training is designed to be accessible without a technical background, while staying relevant for advanced profiles. Everyone walks away with something concrete.
Pricing
An investment
that pays off.
The training is a standalone investment — your team walks away with the knowledge to move forward on their own. Accompaniment builds your architecture. The partnership grows it with you.
Step 1
Base
training
$4,950
per group · up to 12 participants
The 3 full half-days, in person at your offices. Your team walks away with complete understanding of the ecosystem, a first functional agent, and a defined responsible AI governance framework.
- 3 sessions of ~4 hours each
- 3 hands-on workshops in pairs
- Pre-configured workshop environments
- AI governance framework (6 principles)
- In-person training at your office
- 30 days of email support after
- Digital reference guide
Works out to ~$412 per participant for a group of 12. Travel costs extra outside Québec region.
Step 2
Implementation
support
$1,200
per day · custom
We build your AI architecture together. Company Vault, specialized agents, workflows, connection to your tools. You leave with a functional system, not a recommendations report.
- Mapping your existing processes
- Building the Company Vault
- Deploying 2-3 specialized agents
- Connecting to your tools via MCP
- Creating workflows and skills
- Real-conditions validation
Typically 3 to 10 days depending on complexity. Custom quote after evaluation.
Step 3
Ongoing
partnership
Custom
monthly hours block · recurring contract
A block of hours reserved each month to evolve your system. New agents, new workflows, optimization, applied tech watch. You have a dedicated AI advisor.
- Reserved hours each month
- Evolution of agents and workflows
- Integration of new AI capabilities
- Priority support and strategic advice
- Periodic AI governance review
- Applied tech watch
The volume of hours and frequency are defined together based on your needs.
Your instructor
Not coming out
of a classroom.
Someone who spent 10 years building and deploying systems structurally identical to an AI agent.
I'm Louis-Charles Cuierrier, electrical engineering graduate (B.Sc., Université Laval). 15 years in sales total — I started young in B2C — including 10 years in B2B sales at a Québec manufacturer in telecom and public safety, from field technician to VP Sales & Marketing.
I designed, deployed and configured complex monitoring and automation systems for clients across North America. Monitoring is exactly like AI: take a black box, give it rules, triggers, actions. I already think that way.
I built and delivered complete technical training plans for demanding audiences — installation technicians and professional engineers — traveled to clients across North America, and presented to audiences of 100+. Training technical teams — that's my natural ground.
I speak to technicians as easily as to executives. I simplify the complex. And I bring AI as a concrete tool — not as magic.
10
years in B2B sales
3
markets CA · US · MX
100+
people trained live
FR/EN
bilingual
Education
B.Sc. Electrical Engineering — Université Laval, 2017
AI Certification
Lead With AI — MasterClass × Microsoft, May 2026
AI Strategy
Responsible AI
AI Governance
AI Risk Management
AI Leadership
Team AI Adoption
Verify the certificate ↗
Next step
Ready to put AI
in your team's hands?
The first call is about calibrating the training for your team and sector. We look together at whether it's the right moment — and if yes, we plan the 3 sessions.