Ask Claude or ChatGPT about invoicing, customer balances, stock and orders in Cegid PHC, answered from your PHC database at the moment you ask.
Read from Cegid PHC when someone asks. Nothing exported, nothing copied, and changes only where you allow them.
What did we invoice this month by salesperson, compared with the same month last year?
Which customers owe us for more than 60 days, and how much does each one owe?
Which items are below minimum stock in the main warehouse, and who did we last buy them from?
Keep PHC’s documents read-only: invoices, orders and receipts belong in PHC itself, where numbering, totals and certification rules apply. Writes suit your own fields and tables, only approved ones, and never on the User role. Every write is on record.
Under the hood: Cegid PHC keeps its data in SQL Server, so this runs on our SQL Server connection, with the same guards: one statement per call, approved tables only, and nothing beyond the Cegid PHC database.
PHC holds the truth about the business: every invoice, every customer balance, every item in stock. Getting a new report out of it usually means a PHC partner, a custom listing or an Excel export someone repeats every month. Connected once, the people who run sales, purchasing and finance ask in plain language and get the answer from PHC itself, in the AI they already use, next to the CRM, the shop and the inbox.
An admin connects Cegid PHC one time. Nobody hands over a personal login.
Each answer is read at the moment of the question. Nothing is exported or copied, so there is nothing to keep in sync.
What each role can see and do holds in every AI your team uses, and in your own agents.
Every question and every action, by every person and agent, in one audit trail.
Cegid PHC CS, and PHC Evolution where its database runs on a SQL Server you control, on your own servers or hosted. FrontierOps connects to the SQL Server database PHC keeps its data in, which must be reachable from FrontierOps; we work that out with your IT or your PHC partner. PHC GO runs in Cegid’s cloud without database access, so it is not covered today.
No. Nothing is installed in PHC and no Cegid API is involved. It reads the SQL Server database PHC already uses, through a SQL login you create for it (Windows authentication is not supported), limited to the tables you approve.
During setup we map PHC’s tables to plain names: ft is invoicing, cl is customers, st is items, bo is internal dossiers, plus your own user fields. Those definitions guide every answer, so “sales last month” means the same thing for everyone who asks.
PHC’s own assistants work inside PHC. FrontierOps puts PHC in the AI your people already use, Claude, ChatGPT or your own agents, next to the CRM, the shop, email and the rest of your systems, under one set of permissions and one audit trail. The two work side by side.
Yes. Claude and ChatGPT answer in the language they are asked in, so “Quanto faturámos este mês por vendedor?” works as well as the English question.
Yes, and one built for a whole company. FrontierOps gives Claude, ChatGPT and any other MCP client one connection to Cegid PHC, set up once by an admin. What each role can read or change is decided in FrontierOps, and every call is on record, instead of each person running their own server with their own login.
A direct connection links one AI to Cegid PHC through one person’s login, with that person’s settings. That works for one person. For a team it means no shared definitions, no single place to decide who can see or change what, and starting over when you switch AI. Through FrontierOps, Cegid PHC is connected once, for every AI and every agent, under one set of permissions, with every call on record.