AI-enabled manufacturing ERP

Ask questions about orders, products, production and purchasing; the assistant only sees what the user is allowed to see. For how-to questions the assistant searches the manual first and also finds articles when you use another common term, such as trim loss or material surcharge for material loss.

AI tools for orders at risk, orders due soon, stalled orders, bottleneck operations and late purchase orders.

Search by meaning across business data and documents through a vector database, with permission filtering.

Text recognition and extraction from incoming purchase invoices and customer orders.

Have a customer PDF order read by an AI reader that checks order number, dates, quantities and prices against the document; show differences with the sales order and apply the lines. A ribbon button reads in a received customer order directly and the system also recognises by itself which document is a customer order; it learns per customer from earlier imports.

Record, transcribe (also locally with Whisper/Vosk) and generate an agenda and summary.

Support for several providers (Anthropic, OpenAI, Google, Azure) and local models through Ollama.

Model Context Protocol endpoint with tools and resources, secured with API keys, filtered by rights and features, write actions logged in the timeline.