Any format
EDIFACT, flat files, CSV, Excel, email, PDF, free text and voice. Each partner's format is no longer your problem.
AI reads and proposes, rules validate and convert to Sage X3, and every decision is logged. With an MCP server for AI agents.
EDIFACT, flat files, CSV, Excel, email, PDF, free text and voice. Each partner's format is no longer your problem.
Generative AI reads the document and proposes the mapping. Rules validate it, convert the codes to Sage X3 and log every decision. Nothing enters the ERP without a trail.
Ports and logistics are the proof. Distribution, manufacturing, freight forwarding and pharma use the same engine. A new partner is configuration, not a project.
The GANTRY MCP server connects Claude, Copilot or another agent to Sage X3, with each user's own permissions and a log of every request.
In production at a pan-African logistics group, with messages from terminals, shipping lines and customs posted to Sage X3 and the result of every line sent back.
If you recognise two or more, it is worth a conversation. This list is where we start.
Bring EDI and AI to your Sage X3 customers without building custom integrations. We will walk you through the partner model and how we support the first projects.
CODECO, COARRI, COREOR, CUSCAR and carrier manifests come in validated and converted, with the result of every line sent back.
ORDERS, DESADV, INVOIC, DELFOR, IFTMIN and orders in Excel or email, posted to Sage X3 with the same rules and the same log.
Agents ask, explain and act. Rules decide and every request is logged.
“Which containers came in today and are not in Sage X3 yet?”
“Why did Sage X3 reject line 14?”
Submit or reprocess a document, with a person's approval.
Securely: each user's own permissions, separation between companies and a log of every request.
Leave your detailsWe will propose a time for the video call on the next business day.
Send a fileA real example of what you receive: EDIFACT, Excel, PDF or an email. It can be anonymised.
20 minutes on a video callWe show you how that file lands in Sage X3 and what it takes to put it into production.
No. It was born in ports and logistics, where it is in production, but the engine is the same for any message: orders, dispatch advices, invoices, forecasts or transport instructions. Each new partner is a configuration profile.
Not on its own. The AI reads unstructured documents and proposes the mapping; validated rules decide and, when needed, a person does. Every decision is logged, with the source file.
It is the standard way to connect an AI agent to a system. The GANTRY MCP server lets the agent query, explain rejections and submit documents to Sage X3, always with the permissions of whoever is asking.
No. A lightweight agent installed inside the customer's network delivers the files to Sage X3 and returns the result. There are no inbound connections.
Book the 20 minutes and choose “Sage partner” in the form. We will explain the model, the training and how we support the first projects.