How Itsalga scaled orders and production without scaling headcount, on one AI-Native ERP.

Background: A growing producer with lean operations

Itsalga produces artisanal seaweed-based dishes in the Basque Country, made from local, seasonal ingredients and sold under two brands, Itsalga and Coquin, in supermarkets and food retailers across the region.

Operationally, Itsalga manages everything a larger food producer does: retail order intake, production planning, inventory, multi-brand reporting, and invoicing. What it doesn't have is a department for each. That entire cycle sits with Corentin Notte, alongside running the business itself

Challenge: Operations capacity was the ceiling on growth

Food production runs on a weekly rhythm. Retail orders come in, production is planned against available stock, manufacturing orders are issued, deliveries ship, invoices follow. Conventional ERPs are designed around a team, with a user in each module. When one person covers the full cycle, every new customer adds hours of administration to the week.

Itsalga's two-brand catalog compounded the load. Each product carries four levels of structure, from brand down to packaging format, which meant longer order entry and reports that required filtering before they said anything useful. Winning a new retail listing was commercially straightforward and operationally expensive. Growth would have meant hiring for admin, not for production.

"Everything went through me. Every order, every status, every report. The ERP recorded the work, but I was the one doing all of it.": Corentin Notte, Itsalga

Solution: An ERP he can delegate to

Piston is an agentic ERP: Corentin describes the outcome he needs, and Piston carries it out in the system, end to end. Every action lands as a normal ERP record he can open, check, and adjust, so delegating the work never means losing sight of it.

Production planning shows what that changes. Corentin asks Piston to create the manufacturing orders needed to cover confirmed orders, taking current stock into account. Piston reads the order book, checks inventory, works out what actually needs to be produced, and creates the manufacturing orders itself. What used to be an afternoon of cross-checking is now a brief he gives and a plan he reviews.

The same collaboration runs through the week. A retail order such as "enter the order for Leclerc Capbreton, delivery May 7, 3 parcels of each Itsalga reference" is created as written. The kitchen gets its prep sheet consolidated by brand on request. When Itsalga's accountant asked for a stock count of trays, sleeves, and dough, Corentin forwarded her question to Piston nearly word for word and sent back the answer the same day. One question traveled from the accountant through Corentin to Piston and back, and nobody opened a module.

Impact: Room to double volume, no new hires

In a single month, Corentin ran his entire order-to-invoice cycle through Piston: order entry, status updates, manufacturing orders, plus the analysis a larger producer would assign to a dedicated role. Production volumes for April, May, and June. Revenue by brand. Customer segmentation between Coquin and Itsalga. The way he phrases requests tells its own story. One brief simply read, "give me the most complete analysis you can of our customers and products." That is how you talk to a colleague you trust, not how you fill in a form.

The commercial consequence is capacity. Doubling the order book no longer doubles the administrative week, because order entry stays a sentence and planning stays one brief regardless of volume. The next retail listing is a production decision, not a hiring decision.

"I just tell Piston what the week looks like, and it gets done.": Corentin Notte, Itsalga

Small producers rarely stall on demand. They stall when the operations behind each new customer accumulate on the same few people. Itsalga now runs the same operational rhythm as producers many times its size, with the team it already has.