What's been happening.

We are rebuilding Ardesia with the memory at the centre. It reads email and WhatsApp, keeps a page for every customer with the source of every statement, keeps a company memory for what applies to every customer, and turns what it knows into correct quotes. Argo lives inside that memory: it answers with sources, prepares quotes, and proposes changes that a person confirms. It runs on open models on our own servers, so customer data never passes through a third-party model provider. The email reading, the memory and the pricing engine carry over from the previous version; the data model and the interface are new. We are building it with design partners.

Over the summer Ardesia grew toward a full commercial system: invoices, delivery notes, payments, the purchase side. On 16 September we stopped and went back to what the product is for. Ardesia handles quotes and everything before them: what a customer asks, what was offered, what was agreed. Invoicing and the rest of the back office stay with the gestionale that already does them. The fiscal work is frozen, and the entries that described it are no longer in this log.

Every table now filters by the column you are looking at, and says so when a filter is on. Product kits let you pick an assembly once (a belt with its edges, slats and guide) instead of rebuilding it line by line on every order. Customers' own item codes can be recorded and printed on documents.

Ardesia is the product; Silica Systems is the company. We've acquired silica.systems as the domain and reorganised the website to reflect the split: Silica is the publisher, Ardesia is what we publish. The change is purely structural: the team, the product, and the work are all the same. Earlier, we ran everything under the Ardesia name and treated the company and product as one. With a defined scope on Ardesia and a horizon for what may come after, the company needed its own identity.

Prices now live in proper price books. Several companies can share one account, each person can be allowed or not to change prices, and every change is kept in a history. Price books handle more real cases: a price that starts on a given date, a contract rule for one customer, quantity tiers that check themselves, and a floor below which a product does not go.

Started building the marketing site: hero, platform walkthrough, Argo deep-dive, FAQ. The site is being built specifically around the product as it exists today, not as it might exist later. Aimed at office managers in manufacturing companies and the people who decide what software they use.

Spent a day with our first design partner walking through how their office actually runs: every email, every WhatsApp thread, every quote, every gap. Several of the choices Ardesia is built around (the daily ledger format, the customer-scoped Argo, the active-cycle scope) come directly from this conversation.

Two earlier tools: Corridor (office automation with AI) and LexVault (CRM and communication ledger); have been merged into a single product, Ardesia. The CRM and ledger work that lived in LexVault is now the foundation of Ardesia's customer-record view; Corridor's office-automation logic is now Argo, the AI working layer over everything. The merge was a pivot in scope: from "tools that help with parts of the office" to "one product that handles the office's active cycle."

One tool, one workflow, one place. The two previous products will not be developed further as separate offerings.

Ardesia begins as an idea: one workspace for the manufacturing office (invoices, orders, documents, and messages) with an assistant that runs on the customer's terms.

The premise is simple. Office managers in manufacturing companies don't need another tool to add to the stack. They need fewer windows, fewer tabs, and a place where the supplier email, the order it refers to, and the reply that goes back all live together.

From here, the work starts.