RFID, mobile devices, PLM, AI: technology adds nothing if the process it supports is unclear. I help you clarify the need, test the solution in real conditions, then roll it out with the teams.
My approach in 4 steps
- Understand the real work: flows, constraints, team frustrations.
- Define the expected value: starting point, KPIs, success criteria.
- Test in the field: limited pilot, before/after measurements, decision to scale up or not.
- Roll out and embed: adapted process, training, usage monitoring.
Technologies and success factors
| Technology | Business problem | Success factor |
|---|---|---|
| RFID | Long inventories, unreliable stock | Tags and readers tested in the field, inventory process redesigned |
| Mobile devices (mobile SAP) | Stock updated with a delay | Data entered at the moment of the action, simple process for the operator |
| PLM | Multiple document versions, scattered tools | A single platform and clear approval workflows |
| AI documentation assistant | Information scattered across project documentation | Up-to-date documentation and a clear scope |
| AI planning agents | Repetitive decisions, high data volumes | Reliable master data first |
Track record
- Microstore (2025-2026): RFID inventory. Under 30 minutes instead of 2 hours, with a team of 2 instead of four teams.
- Sanofi (2024-2025): AI documentation assistant in production on a Master Data programme.
- McPhy (2022-2024): 3DEXPERIENCE PLM in 3 countries. A single platform, and everyone knows which version to work on.
- Solvay (2012-2018): mobile SAP across several plants for real-time inventory management.
Frequently asked questions
Should you start with a pilot?
Yes, whenever the technology is new to the company. A short pilot with before/after measurements avoids scaling up a solution that does not hold up in the field.
Can AI fix our data?
No. It can help spot anomalies, but above all it amplifies what it is given. Governed data remains the prerequisite.
How do you get a new tool adopted?
By building it with users, by simplifying their tasks rather than adding to them, and by measuring actual usage after go-live.
Let’s talk about your project
I am Jérôme Dumas, an independent supply chain and IT consultant based in Lyon, France, working on assignments across France and Europe.
Do you have a process to improve or a technology to assess? Tell me about your context and challenges, and we will see together whether I can help.