Customer job specs, vendor invoices and CVs all land in the same mailboxes. Varg reads each one, including its attachments, decides what it is, pulls out the fields your team would re-key, and sends it to the right desk.
Someone reads every mail to decide whether it's a customer brief, a supplier bill or a CV, then forwards it. Urgent quotes wait behind newsletters.
Quantity, size, pages, paper, colours and binding are copied from the customer's email into the estimating sheet, one field at a time.
Vendor bills arrive as PDFs and phone photos in personal inboxes, reach accounts late, and miss their due dates.
Connects to the mailbox you already use and picks up every new mail with its attachments. Handled mail is labelled so nothing is read twice.
An AI model reads the sender, subject, body and the attached PDFs or images, then decides what the mail is, with a confidence score and a one-line reason.
Pulls the fields that matter for that type: job specs from a brief, amounts and PO numbers from an invoice, experience and skills from a CV.
Files it in the right team's queue and keeps a record of every decision. Anything below a 70% confidence threshold goes to a person for review.
Real output from the POC on sample print-industry mails. Pick one. In the POC each mail lands in its team's tab on the dashboard.
Sample senders and documents are synthetic test data. The classification, reason and fields are the engine's unedited output.
OCR plus AI extraction, PO matching, exception routing and ERP sync. Invoices sorted here flow straight into it.
Record a process on the shop floor once, get a versioned SOP with screenshots, roles and compliance references.
Emails, PDFs and job briefs sorted, extracted and routed to the right team automatically.
Open the demo and paste any enquiry, invoice or CV, or connect a test mailbox and watch it sort itself.