Case Studies
Trusted by these teams.
Data & AI projects delivered, operated for the long run, with gains measured case by case.
Data & AI projects delivered, operated for the long run, with gains measured case by case.
The team is building an alternative map to the Michelin Guide, dedicated to venues accessible to people with disabilities. Every city requires exhaustive scouting, then field surveys through an observation questionnaire.
A redesigned observation questionnaire built for structured tracking, scrapers that automatically pull the venue list for each new city, a single database connected to an online reporting dashboard. The entire project now fits on one screen.
Engagement ran through the early phase of the project. The system now runs autonomously and scales without permanent technical reinforcement. Financial gain estimated at a fully-loaded rate of €45/h.
Maeva & Co runs a portfolio of rental properties through its PARCEL operation. Every week, the useful data, bookings, revenue, extras, nights, site traffic, lived scattered across Excel exports, emails and separate spreadsheets. Everything relied on manual re-entry.
A system that triggers reports on its own, downloads exports with no intervention, parses the Excel files, calculates revenue, extras and nights, pulls site traffic, then feeds it all into the right place in Google Sheets with status tracking. Géraldine opens her spreadsheet: everything is already consolidated.
The system runs autonomously. No more exports to open, no more copy-pasting: a single spreadsheet, always current, to make decisions from. Financial gain estimated at a fully-loaded rate of €45/h.
Training program to make Groupe SEB's teams self-sufficient on AI: getting hands-on with LLMs, building specialized chatbots for their use cases, and scoping automatable tasks with an AI feasibility rating for each.
Hands-on training built on their real use cases: getting up to speed with LLMs, building specialized chatbots, and listing tasks with an AI feasibility rating to prioritize what's worth it. Teams leave with the method and assets they can use immediately.
Ongoing engagement. Goal: teams self-sufficient on AI, able to identify, scope and prioritize their own use cases.
Across our production cases: over 35 hours a month recovered at Michelin Group, over 8 hours a week at Maeva & Co, roughly 18 to 19 k€ a year in recovered cost per project, calculated from time saved valued at fully-loaded hourly cost.
We trigger and pull the exports automatically (Amenitiz-type), parse the Excel files, calculate revenue, extras and nights, then consolidate everything into a single table updated every 48 hours, with no manual re-entry.
We start from hours actually saved, multiplied by the fully-loaded hourly cost of the role involved. Example : 8 hours a week saved at 45 € fully-loaded is about 18 k€ a year, before counting gains in reliability and faster decisions.
Yes. For Groupe SEB : getting hands-on with LLMs, building specialized chatbots on their real use cases, and listing tasks with an AI feasibility rating to prioritize what's worth it. Teams leave self-sufficient.
No. We always start with a 20-minute call to understand your context and confirm it's relevant. The priced audit only comes after, only if it's useful, no commitment.
Industry and large groups (Michelin, Groupe SEB), rental management and services (Maeva & Co), in data engineering, automation and AI. Engagements run in France and across Europe.
We always start with a 20-minute call to understand your context and what's at stake. Together we see if it's the right fit, and scope what's next only if it makes sense. No commitment, nothing imposed.