FREN
Preview · Your 1:1 advisory session with Rita Alaoui

Your advisory session: Introduction to Agentic AI.

A preview of your 1:1 advisory session. Here you see the structure; in the session, we go through it together, applied to your organisation, your teams and your goals. You also leave with a notebook to capture your takeaways and next steps.

Book a 30-min call →
Inspirational trAInings · Workbook preview01
Our session

Today's agenda.

Inspirational trAInings · Workbook preview02
We cover this together in your session
AI fundamentals

From AI to agents, one layer at a time.

Every agent is AI, but not all AI is an agent.

Artificial Intelligence
Machine Learning
Generative AI
Agentic AI
Artificial Intelligence
Machine Learning
Generative AI
Agentic AI
Remember
Inspirational trAInings · Workbook preview03
We cover this together in your session
AI fundamentals

Key AI vocabulary.

Ten terms you’ll hear in every AI conversation. Shared vocabulary is the fastest way to align a leadership team.

LLM

Prompt

Multimodal

Token

Hallucination

Sycophancy

RAG

Fine-tuning

Guardrails

Human in the loop

Inspirational trAInings · Workbook preview04
We cover this together in your session
AI fundamentals

Three levels of AI integration.

Prompting, workflows, AI agents: when and how to use each.

Level 1

AI tools

Operator

Level 2

AI workflows

Designer

Level 3

AI agents

Governor

Inspirational trAInings · Workbook preview05
We cover this together in your session
Prompting

Prompt with design and technique.

Five ingredients, the CRAFT method, turn a vague request into a result you can use.

C

Context

R

Role

A

Action

F

Format

T

Test

Example · announcing an AI usage policyContext
Role
Action
Format
Test
Inspirational trAInings · Workbook preview06
We cover this together in your session
Prompting

Create your prompt library.

Pick three tasks you do every week. We write the prompts together, then you save the best versions where your team can reuse them.

1Your prompt #1

C
R
A
F
T

2Your prompt #2

C
R
A
F
T

3Your prompt #3

C
R
A
F
T
Inspirational trAInings · Workbook preview07
We cover this together in your session
Workflows

AI workflows: same steps, every time.

A fixed sequence of steps, some done by AI, that runs the same way every time. You design the path; AI does the work at each stop.

Trigger
→
AI step
→
Rule
→
AI step
→
Human

A workflow is a train 🚆

An agent is a taxi 🚕

Inspirational trAInings · Workbook preview08
We cover this together in your session
Agentic AI

The basics of an agent.

Strip any agent down and you find three parts working together.

AGENTachieves a goal

Model

The brain

Loop

The project manager

Tools

The hands

Definition
Inspirational trAInings · Workbook preview09
We cover this together in your session
Agentic AI

Six steps to build an agent.

It's a loop, not a line. We walk through each step with one of your own processes.

iterate monitor over time DEFINEDATADESIGNEMPOWERPROTECTTEST &DEPLOY
Inspirational trAInings · Workbook preview10
We cover this together in your session
Agentic AI · Step 1/6

Step 1: Define.

Start with the problem, not the technology.

iterate monitor over time DEFINEDATADESIGNEMPOWERPROTECTTEST &DEPLOY
What it means
Questions to ask
Example · customer emails
Inspirational trAInings · Workbook preview11
We cover this together in your session
Agentic AI · Step 2/6

Step 2: Data.

No data, no agent. Garbage in, garbage out.

iterate monitor over time DEFINEDATADESIGNEMPOWERPROTECTTEST &DEPLOY
What it means
Questions to ask
Example · customer emails
Inspirational trAInings · Workbook preview12
We cover this together in your session
Data quality

Six properties of good data quality.

Garbage in, garbage out.

Completeness

Uniqueness

Timeliness

Validity

Accuracy

Consistency

Inspirational trAInings · Workbook preview13
We cover this together in your session
Data first

Apply your business expertise to think like a data engineer.

Pick one AI problem from your matrix, then map the data it needs, level by level.

The AI problem we want to solve
Data sources
1st level

Internal data

CRM, ERP, spreadsheets, emails, documents

2nd level

External data

Open data, market studies, partner and supplier data

3rd level

Think outside the box

Industry data, third-party APIs, public datasets

Inspirational trAInings · Workbook preview14
We cover this together in your session
Agentic AI · Step 3/6

Step 3: Design.

Draw it on paper before anyone builds it.

iterate monitor over time DEFINEDATADESIGNEMPOWERPROTECTTEST &DEPLOY
What it means
Questions to ask
Example · customer emails
Inspirational trAInings · Workbook preview15
We cover this together in your session
Agentic AI · Step 4/6

Step 4: Empower.

Treat it like a new hire: a clear brief, and only the access it needs.

iterate monitor over time DEFINEDATADESIGNEMPOWERPROTECTTEST &DEPLOY
What it means
Questions to ask
Example · customer emails
Inspirational trAInings · Workbook preview16
We cover this together in your session
Agentic AI · Step 5/6

Step 5: Protect.

Decide where it must stop before it ever starts.

iterate monitor over time DEFINEDATADESIGNEMPOWERPROTECTTEST &DEPLOY
What it means
Questions to ask
Example · customer emails
Inspirational trAInings · Workbook preview17
We cover this together in your session
Agentic AI · Step 6/6

Step 6: Test & Deploy.

Launch small, measure, then scale. And never leave it without an owner.

iterate monitor over time DEFINEDATADESIGNEMPOWERPROTECTTEST &DEPLOY
What it means
Questions to ask
Example · customer emails
Inspirational trAInings · Workbook preview18
We cover this together in your session
Design the agent

The six parts of every agent.

Whatever the platform, every agent is built from the same six parts. We design yours, part by part.

Model

Instructions

Tools

Memory

Guardrails

Loop

Inspirational trAInings · Workbook preview19
We cover this together in your session
Governance

Governance best practices.

Treat an agent like a new colleague: a clear scope, limited access, and someone in charge.

Human in the loop

Data privacy & security

Who trains & monitors it?

What if it goes wrong?

Ownership

Where do we set the bar?

Bringing people along

EU AI Act compliance

Inspirational trAInings · Workbook preview20
We cover this together in your session
Governance

The EU AI Act: risk decides the rules.

The higher the risk to people, the stronger the obligations. We place each of your use cases on the pyramid.

Unacceptable · banned
High · strict obligations
Limited · transparency
Minimal · no specific obligations
Key dates
Feb 2025
Aug 2025
Aug 2026
Dec 2027
Aug 2028

Simplified overview, not legal advice. Check with your legal or compliance team.

Inspirational trAInings · Workbook preview21
We cover this together in your session
Governance

Data security and sovereignty.

Five pillars to protect your data.

C

Confidentiality

Only authorised eyes.

I

Integrity

Data stays unaltered.

A

Availability

Reachable when needed.

C

Control

You keep control of your data.

C

Compliance

You meet the regulations.

Inspirational trAInings · Workbook preview22
We cover this together in your session
Governance

AI Financial Governance.

Keep AI costs under control.

Who manages costs, together
Finance+Tech+Business

People

Accountability, skills

Process

Budgets, reviews

Technology

Billing tools, automation

Inspirational trAInings · Workbook preview23
We cover this together in your session
Brainstorm & prioritise

Your AI use-case matrix.

We list your ideas, then place each one by impact and effort, to find the right first project.

Scope of impact

Disruptions

Transformations

Quick wins

Developments

1
2
3
4
5
6
7
8
Complexity / time to implement
Your use cases
1
2
3
4
5
6
7
8
Inspirational trAInings · Workbook preview24
We cover this together in your session
Put it all together

Your agent brief, on one page.

Pick one use case from your matrix and answer six questions: that’s your brief, whoever builds it.

Our use case

1Trigger

2Data

3Outcome

4Boundaries

5Ownership

6Output

Inspirational trAInings · Workbook preview25
We cover this together in your session
Your action plan

The next 90 days.

You leave with concrete actions, owners and a way to measure success.

30 days

Learn & pick

60 days

Pilot

90 days

Measure & scale

Inspirational trAInings · Workbook preview26
Keep going

Notes & next steps.

My key takeaways
Questions for our follow-up

In a 2-hour 1:1 advisory session, we go through every slide together, for your organisation and your goals. You leave with a notebook to capture your takeaways, your use cases and your next steps.

Book a 30-min call →

Rita Alaoui · Inspirational trAInings · inspirationaltrainings.com

Inspirational trAInings · Workbook preview27