AI integration case studies

Advisory, training, marketing, AI agents and automations. Every project starts from how your team actually works.

Process mapping and identification of AI use cases

AI advisory

Mapping processes to find AI use cases

Landscaping company

24 tasks sorted, 4 quick wins underway

View the project
Automating job site preparation from a site survey video

AI advisory

Automating job site preparation

Landscaping company

The same information entered once instead of four times

View the project
Before, all I knew how to do was write a prompt. Today I produce landing pages, guidebooks, carousels and presentations. Claude is part of the team now.
Website redesign with Claude Code

Design

Website redesign with Claude Code

Qiplim

3 days, 0 lines coded by hand, Lighthouse 100

View the project
AI marketing copilot driven by Claude Code

Internal

AI marketing copilot

Internal project

12 features, 8 specialized modules, open-source template

View the project
AI graphic design studio

Design

Horizon Studio: an AI visual creation app

Horizon Humain

100+ custom images generated with Gemini 3 Pro

View the project

TRAINING

Claude for marketing

Three deliverables built from your own materials, and the method to make them again without help.

See the training
Service portfolio design

Design

ServiceNow & ESM projects book

N2 Help & Solutions

8 case studies, 43+ prospects, done in 3 days

View the project
Automated AI prospecting agent

Marketing

Automated AI prospecting agent

Digital services company

4 qualified meetings in the first month

View the project
AI use case identification workshop

AI advisory

AI use case identification workshop

Quebec notarial firm

~CAD 150,000 a year in estimated savings

View the project
What struck us was the way he introduced us to AI: with enthusiasm and concrete examples from his own practice, he made the subject accessible and inspiring. We adopted these tools naturally and with confidence.
AI agents and automation training

AI training

AI agents and automation training

Tech company

35 people trained, AI strategy formalized

View the project
Go-to-market strategy for an AI SaaS platform

Marketing

Go-to-market strategy for an AI SaaS platform

Qiplim

210 users, €4.77 per lead

View the project
Microsoft Copilot training

AI training

Microsoft Copilot training

SMB

55 people trained, 8.5/10 satisfaction

View the project

TRAINING

Generative AI and Microsoft Copilot

One day to get your team prompting methodically, with the right security habits.

See the training
With strong organizational skills and a perfect command of new AI tools, Jessy does excellent work for Mutuelle Entrenous in just 1 day a week. Quality and productivity are there!
Website redesign project management

Design

Website redesign project management

N2 Help & Solutions

Engagement from 66 to 90%, custom-built website

View the project
E-learning course on growing your business with AI

AI training

E-learning course “Growing your business with AI”

Mon Parcours en Ligne

21 hours, 472 slides, 14 hours of video

View the project

Have a project in mind? Let's talk

Book a discovery call

30 minutes, no commitment.

The STEP method in the field

All my projects rely on the STEP method developed by Paul Leonardi (UC Santa Barbara): Segmentation, Transfer, Education, Performance. On an engagement, each step produces concrete deliverables.

Segmentation

Map the tasks

One-on-one workshops with each member of the marketing team to sort their tasks into three categories: augment with AI, automate, keep for humans. Output: a shared view of the scope and the priorities.

Task matrix per employee + prioritized list of 5 to 10 use cases

Transfer

Evolve the roles

The time freed up by AI assistants and agents is reinvested in higher-value tasks. I produce a revised job description for each affected role and a plan that explains to each employee what changes for them in practice.

“Before and after AI” job descriptions + role evolution plan

Education

Anchor the learning

A path tailored to each profile, from a half-day introduction to the RS7344 certification [French official certification code] (21 hours). I deliver the materials, the shared prompt library, the internal AI use policy and the team rituals to keep learning on the job.

Learning path + prompt library + AI use policy

Performance

Measure the ROI

Indicators set during scoping: time saved, real adoption, output quality, business impact. I set up the dashboard and the quarterly review cycles to adjust over time, with figures that hold up in front of the executive committee.

AI KPI dashboard + quarterly steering committee

For the theory behind the method, read the full article on STEP.

Questions?

Engagements range from a 3-person small business to marketing teams of 10 to 15 people at large corporations. Size matters less than a clear need and leadership's commitment to the exact scope of the engagement.
I run AI projects for marketing teams myself: that's my core expertise. For other business functions (sales, HR, finance, operations, tech), I work through my partnership with Pando Studio, which brings in specialized AI experts on demand to handle AI training and integration for any department.
From 1 day for a scoping workshop to 6 months for a full rollout with training. Most engagements last 4 to 12 weeks, with an operational pilot in 4 to 8 weeks before scaling up.
Both, depending on your constraints. I'm based in Nantes, France, and regularly work in Paris, Lyon and Clermont-Ferrand, as well as abroad (Quebec, Morocco, etc.). Scoping sessions and workshops work just as well remotely as on site: the facilitation methods adapt to the size of the team, the format and the organization's constraints.
Three formats, detailed in the offers section: €450 excl. VAT per day per person for a catalog training course in an open session, from €5,900 to set up the AI Marketing Cockpit for your team (audit, configuration, connected tools, AI use policy and training), and a custom quote for an advisory engagement or a custom business agent. Pricing follows a first 30-minute call.
I take on engagements where the impact is measurable, I can get access to the teams doing the work, and decisions are quick. I turn down vague or purely cosmetic projects, and those where leadership isn't truly committed to the change.
The work adapts to your stack. On recent projects: Claude Code, Microsoft Copilot, Gemini, OpenAI, n8n, Make, Brevo. The choice is made during scoping based on your constraints (tools approved by IT, GDPR, budget). My full selection is in the article on the 5 AI tools I recommend in 2026.