📘 Corporate AI training: our free roadmap for a successful rollout
Download our AI roadmap
★★★★★
4,9/5 sur Google
🔌 Connect Claude to Pennylane: our step-by-step setup guide
Download the Claude x Pennylane guide★★★★★
4,9/5 sur Google
.svg)
📘 Corporate AI training: our free roadmap for a successful rollout
Download our AI roadmap★★★★★
4,9/5 sur Google
🔌 Connect Claude to Pennylane: our step-by-step setup guide
Download the Claude x Pennylane guide★★★★★
4,9/5 sur Google
The AI Sisters Lab scoped the AI projects of JCDecaux’s IT department in France with a quantified ROI for each one: five use cases scored, two MVPs selected on their estimated gains.
in estimated annual gains across all cases
use cases analysed, scored and costed
from kick-off to findings
MVPs selected, with specifications and roadmap
“
You went beyond all our expectations.
Eric Hahn, Head of Skills, JCDecaux
Before
Each department pushed its own, with no shared criterion to decide.
With the AI Sisters Lab
Five cases scored on the same grid, and one case dropped on the figures: €7K of gains for €89K of investment.
Before
A time measured with the teams during the co-design workshop.
MVP target
For around €686K of estimated annual gains, with a POC on real data before the go decision.
From 15 December 2025 to 4 February 2026, the AI Sisters Lab ran an AI use-case scoping and prioritisation mission for JCDecaux's French IT department.
For each case: a questionnaire, a co-design workshop with the teams involved, a combined technical and business analysis, then an impact and feasibility score.
The goal: come away with a 2026 AI roadmap and two MVPs ready to be developed.
In July 2025, AI Sisters had trained the French IT department's teams over two days. The logical next step: decide where to invest. Several departments each had their own AI project idea, and none had been costed.
The challenge for AI Sisters: identify use cases with a high return on investment, technically realistic and genuinely useful to the product, tech and sales teams.
Each department was pushing its own project. A common scoring grid was needed, announced before the workshops.
Real volumes and time spent, gathered with the teams, rather than desk-based assumptions.
Each case had to be assessed within JCDecaux's technical environment, with its access and security constraints.
A short format per use case, so business teams were not tied up for weeks.
Several AI project ideas, none of them costed?
We review them in detail with your teams and give you a well-argued decision.
A Lab team of AI consultants and engineers, supervised by our CTO. The same sequence for every use case, and a weekly check-in with the sponsors.
On 15 December 2025, an alignment meeting with the technical and business sponsors. The process was validated: short interview, questionnaire reviewed by JCDecaux, two-hour workshop, shared summary.
From December to January: UX/UI AI, document AI, sales AI, finance AI. In each workshop: the current process, pain points, volumes and actual time spent.
On 4 February 2026, for each case: a technical analysis and a business analysis, an impact and feasibility score out of ten, an estimate of gains and implementation cost, then a verdict.
On 19 February 2026, for the two cases chosen by JCDecaux: functional and technical specifications, backlog, demo scenario and development roadmap.
Have you chosen your use cases?
Discover our AI POC and MVP support, tested on your data and validated by your teams.
Five use cases assessed against the same grid. Scores are out of ten; gains are estimates built with the teams, to be measured after deployment.
→ Sales AI: impact 8.0, feasibility 7.3
→ Financial statements AI: impact 8.5, feasibility 5.8
→ Sales AI target: from 2 h 35 to 25 min per file
→ Document AI, 360 view: impact 8.4, feasibility 5.8
→ Technical constraints to resolve before committing to development
→ UX/UI AI, mock-ups: impact 5.8, feasibility 9.2
→ Customer visit route AI: impact 3.6, feasibility 7.5
The French IT department came away with a well-argued decision: two MVPs selected, sales AI and financial statements AI, each with its functional and technical specifications, backlog and roadmap.
IT, product, sales, finance: every department has its use case.
Let's talk about the one that will pay off most for you.
Got it. The Lab team will email you to set up a first call.
Your email didn't go through. Check the address, or write to us at contact@aisisters.ai.
A Lab expert will reply to scope your AI agent project, with no commitment.
Data security and privacy
Every Lab project takes your security constraints into account from the scoping stage. Tools are built in your environment or on services covered by a data processing agreement (DPA): your data is not used to train models. Processing complies with GDPR and the EU AI Act.
A free 45-minute scoping session with a Lab expert.
Got it. The Lab team will email you to set up a first call.
Your email didn't go through. Check the address, or write to us at contact@aisisters.ai.
A Lab expert will reply to scope your AI agent project, with no commitment.
“Great training and a very practical audit: we are already seeing tangible gains.”
“AI Sisters supported 150 Webedia employees with truly exceptional shows and workshops. Extremely hands-on, and the productivity gains are already very measurable.”
“Two brilliant training sessions with AI Sisters. Titouan was excellent: highly personalised, relaxed and accessible. The team demystified a subject that seemed so complex. Thank you!”
“It was really great: AI Sisters supported the entire IT department with immediate impact. I recommend them 100%.”
“An unforgettable experience: dynamic, practical and directly useful day to day. The teams are still talking about it!”
“A very dynamic, practical day, perfectly suited to our challenges. The teams came away with tools and ideas to put into practice the very next day.”
A few details to help us prepare your tailored Show. Our team will get back to you quickly.
Phone: +33 7 80 96 07 26
Address: 255 Boulevard Pereire, 75017 Paris
Email: contact@aisisters.ai
.avif)
