JCDecaux x AI Sisters: how the IT department activated the potential of generative AI

Summer 2025 marks a turning point at JCDecaux: for 2 days, the tech and product teams plunged into the world of generative artificial intelligence with AI Sisters. An intensive program of AI training, practical workshops and real use cases, designed to make artificial intelligence And of theAI an immediate operational lever.

A striking opening: the AI Show for all employees!

‍

To launch this immersion, an IA Show brought together all the participants around an acculturation to LLMs (Large Language Models).
THEAI lens was clear: lay the foundations, create a common base of understanding, and above all... inspire through real, professional and transversal use cases.
‍

🔍 Topics covered:

• how artificial intelligence And the Machine learning transform tech, HR, finance or product jobs.
• Generative AI vs. classic AI : what do you really need to understand in a AI training.
• Interactive demos: image, text, automation, agents AI.
• Responsible practices, governance and skills development through the AI training.

A strong, educational and engaging time, to align teams on the concrete challenges of generative AI and show that each AI training can trigger immediate gains.

‍

Workshops 2 & 3: artificial intelligence at the service of developers and POs

The afternoons were devoted to workshops differentiated according to the profiles. Each workshop illustrated how artificial intelligence (AI) And a AI training well-designed can transform daily practices and increase productivity.

‍

Workshop 2 — Developers: JetBrains, ChatJCD, and Code Augmented with Deep Learning

‍

Objective : accelerate delivery, make code reliable, automate daily tech thanks toAI.

🛠 Use cases addressed:

• Refacto smart in JetBrains AI.
• Automatic detection of technical debt with AI.
• Generation of unit tests from functional specs via AI.
• Git push + PR assisted by AI.
• Creation of internal LLM agents (via ChatJCD).

These uses demonstrate that the AI training applied to the code is an asset for all developers.

‍

Workshop 3 — PO & PM: transforming a brief into a usable user story thanks to generative AI

‍

Objective : save product teams time by structuring needs with The AI.
‍

🔄 Steps worked on:

• Reformulation of a vague need (e.g. automation of commercial discounts) thanks toAI.
• Writing a clear user story with functional hypotheses assisted by AI.
• Creating a structured Jira ticket using the AI training on writing specs.
• Definition of acceptance tests with the help of an agent AI.
• Summary of the specification for non-tech stakeholders.

Central exercise:
‍
Digitization of commercial discounts in GALIA with generation AI tickets, tests, MVP roadmap → industrialization possible. One AI training dedicated to this use case made it possible to understand the power of these tools.

‍

Workshops 4: co-construction of use cases by mixed teams in machine learning

The latest workshops mobilized participants to identify and formalize their own use cases. AI, according to JCDecaux challenges and business needs. The co-construction approach showed that each AI training makes it possible to align the fundamentals of artificial intelligence (AI) to concrete problems.

💼 Use cases from the JCDecaux teams (workshop 4)

All the use cases below were proposed by the participants themselves, in connection with their business realities and their daily lives, illustrating the importance of a AI training adapted.

1. AI agent connected to business tools and social networks

• Objective: to access project information more quickly (Jira, GitLab, Confluence).
• Interface: chatbot on Slack/Teams/Intranet.
• Expected result: -30% in lost time, better responsiveness, better use of tools AI.

2. Gherkin automatic test generator for AI training

• Objective: generate standardized acceptance tests from Jira with theAI.
• Implementation: “generate tests” button → LLM → .feature file.
• Expected result: reduction in regressions, better test coverage and increased skills via the AI training.

3. AI assistant for marketing pitches with deep learning

• Objective: create consistent supports from scattered docs using theAI.
• Implementation: chat AI + upload docs → speech summary.
• Impact: + speed, + consistency, + reuse for marketing content. One AI training in digital communication has made the approach more reliable.

4. AI dashboard for DSI projects and artificial intelligence

• Objective: centralize all DSI project data thanks to theAI.
• Features: milestones, risks, Gantt, alerts, automated PDF report generation with AI.
• Impact: better planning, less manual reporting, unified vision.
  ‍

In summary: what JCDecaux activated thanks to AI and machine learning

✔️ Concrete gains in Product & Dev flows.
✔️ Use cases that can be immediately activated in GALIA, Jira, GitLab thanks to theAI.
✔️ An increase in AI skills for all profiles, from freelancers to face-to-face, thanks to AI training.
✔️ A strategic alignment of the IT department on generative AI And the Deep learning, supported by a strategy of AI training continue.

‍

‍

🤝 Tailor-made support by AI Sisters in artificial intelligence

At AI Sisters, our approach is simple: using our customers' tools and contexts to build AI training 100% operational. With JCDecaux, each workshop has been designed to anchor artificial intelligence (AI) And theAI in daily life, taking into account real needs, challenges and opportunities.

We work face-to-face or remotely, with an approach focused on productivity and digital transformation. Teams benefit from tailor-made support, whether for marketing, development, engineering or voice recognition.

THEAI lens remains clear: give businesses the skills, practices, and steps needed to take advantage of emerging technologies, improve their processes, and enroll the AI training as a strategic lever at the heart of daily work.

‍

JCDecaux and AI Sisters are transforming the IT department through AI training and generative artificial intelligence. Concrete workshops to activate AI on a daily basis.