Ten times the peak. The same small team.

JDM Products distributes consumer electronics and lifestyle brands into UK and Irish retail, including Boots. Founded in 1999, it manages more than 3,000 active trading accounts with a deliberately lean team.

Every customer email came into a shared inbox. Someone read it and worked out whether it was a tracking question, an investigation, a return or a refund. Then they opened the right carrier portal, matched the order number by hand and wrote a reply.

At peak season the volume rose three to four times. The team did not. Warranty terms, retailer return rules and brand templates lived in people’s heads and old emails. Internal logistics messages sat in the same inbox as real customer issues.

The team works in Outlook, as before. Each query is also a card on a dashboard, with its full conversation history and a carrier breakdown, moving along a pipeline from new to resolved.

Nothing goes out without a person. The system reads, checks and drafts, and the team decides. Since going live in January 2026 we have held a weekly session with the team. The dashboard was rebuilt on their feedback in February, and 47 corrections from their edits were applied in the first weeks.

In February, JDM’s Commercial Director laid out new AI use cases across three business units, unprompted:

• field-sales briefings
• cross-system search
• new-brand onboarding
• lead detection for its 3PL arm
• SOP documentation

He also asked for a presentation to give to his own board, and he now calls the system a “locally hosted AI team”. The foundation (classification, knowledge base, carrier integrations and the learning loop) is reused by each new workflow.