The context
To supply dozens of job sites across Eastern Canada, Fournier runs a hub-and-spoke model: a large share of orders converges on the Val-d’Or warehouse before being reshipped to the job sites. Hundreds of orders a day, suppliers everywhere, an internal fleet for the transits.
The problem
Order and receiving follow-up happened by email and phone. Orders got lost, lead times stretched, and visibility ended the moment a supplier shipped: material arriving at the wrong job site, no view of what was in transit, job sites waiting for material with no idea when it would arrive. Receptions done directly on site cut the thread entirely: a paper packing slip in hand, nothing in the system — no one knew whether the order had actually been received.
What we built
An application layer, integrated with the existing ERP, that ties the whole chain together: requisitions, supplier orders, warehouse receiving, reshipment and transit between locations, and receiving on remote job sites.
On site, the person receiving takes a photo of the packing slip — AI does the rest: matched against the purchase order, the reception entered into the system, the right people notified.
At the Val-d’Or warehouse, an AI agent picks up incoming orders and groups them for reshipment to the job sites, drawing on a wide range of data sources: it knows what is urgent, where each order needs to go and when. Warehouse clerks only have to handle the loading.
The platform also plugs into the internal fleet’s GPS — real-time location of every transit, more reliable delivery estimates. And AI monitors open orders and follows up with suppliers on its own, taking the chasing off the buyers’ plates.
The results
- 100% of orders traceable, from requisition to on-site reception
- Real-time visibility on everything in transit
- Procurement lead times cut by 40%
- Dozens of hours a week recovered from email and phone follow-ups
- Supplier follow-ups automated — buyers focused on what matters