How Fromental built an AI-native operation for commercial ops

The work sat in four companies

Fromental makes luxury wallcoverings. Trade buyers, sample orders, and four companies sharing a commercial desk, and too much of that desk was export, copy, and re-key.

The commercial team was pulling historic CSVs out of Bridge and building pivot tables to keep a dashboard current. Invoices had to be routed across four companies. Sample requests and Pipedrive deals did not meet unless someone walked them there.

The stack they already had

We did not put in a new system of record. We put the work on Pipedrive, Shopify, Klaviyo, and Bridge, the tools the team already opened every morning.

The sales intel dashboard reads Bridge directly, so the export-and-pivot morning is gone. Invoices find the right company. Enquiries and samples land where a rep can act.

The morning stays with the buyer

25+ workflows are live. Invoice processing is faster. The team gets 200+ minutes back every week, and the outsourced bookkeeping that used to move those bills is gone.

People still decide what to send a Grade 1 client. They do not spend the morning preparing the list.

That is the difference on a commercial desk: the report is current because the data already is, not because someone rebuilt it from a CSV.

Route. Brief. Review.

  1. 01

    Queue the work

    Invoices, samples, and deals as they arrive.

  2. 02

    Route the invoice

    The right company, without re-keying.

  3. 03

    Brief the desk

    Sales intel from Bridge, not a CSV.

  4. 04

    Review exceptions

    People still decide. The list is already there.

What it runs on

PipedriveShopifyKlaviyoBridgePipedriveShopifyKlaviyoBridgePipedriveShopifyKlaviyoBridgePipedriveShopifyKlaviyoBridge

The report is current because the data already is

Before

Manual

CSV exports, pivots, outsourced books

After

5× faster

Invoice routing across four companies

Automation

Have a similar programme?Let's talk.

Tell us the desk and the bottleneck. We'll map it to a workflow, or tell you if we shouldn't.