# How to automate email into MRPeasy > Source: https://netray.co/integrations/mrpeasy-email-automation > Publisher: Netray (netray.co), AI and integration engineering for ERP and manufacturing systems. > Section: Integrations ## Short answer Read a shared mailbox through IMAP or the Microsoft Graph or Gmail API, classify each message, extract the lines with a model, match customers and items against MRPeasy, and create draft records through the REST API for a person to confirm. Start with one message type. ## System MRPeasy ## Capability email-automation ## Intro A lot of MRPeasy work starts in an inbox. A customer sends a purchase order as a PDF, a supplier replies with a confirmation, someone requests a quote in a paragraph of text. A clerk retypes it into customer orders, quotes or purchase orders, and the time adds up in small companies without a back office. Email automation here means reading those messages, understanding them and preparing the MRPeasy record, not sending the order into the system blindly. The goal is that a person reviews a ready draft instead of typing from scratch. This page walks through the pieces, the matching problems that dominate the effort and how to keep it safe. ## Use Cases ### Customer PO to draft order **Description:** Extract customer, PO number, lines, quantities and requested dates from an attached PO and prepare a customer order. ### RFQ to quote **Description:** Read a request for quote, match items to the MRPeasy catalogue and prepare a quote with current prices for review. ### Supplier confirmation to PO update **Description:** Parse an order acknowledgement and propose an updated expected date on the matching purchase order. ### Shipping notice matching **Description:** Match an advance ship notice or tracking email to an open purchase order and note the carrier and ETA. ### Invoice intake **Description:** Route supplier invoices from the accounts mailbox to the right folder and flag mismatches against received purchase orders. ### Customer change requests **Description:** Detect emails asking to change date or quantity on an open order and queue them for the planner with the order attached. ## System Surface - MRPeasy REST API with key and secret for reading items, partners, customer orders, quotes and purchase orders and creating or updating them; verify exact write support by version. - Partner records give email domains and contact addresses to identify the sender; item records give codes and descriptions for matching. - MRPeasy has its own email sending for documents, but no general inbound email to order feature to depend on; inbound parsing is your layer. - Attachments stay in your mailbox or a store of your choice; the API may allow files to be attached to records, so check the current documentation. - SaaS only, so the processor reads mail and calls MRPeasy; nothing runs inside MRPeasy. ## How It Works ### Choose the mailbox access method **Detail:** Use Microsoft Graph or the Gmail API with a dedicated shared mailbox, or IMAP with an app password. Process by label or folder so you never touch personal mail. ### Classify the message **Detail:** Decide if it is a PO, RFQ, confirmation, invoice or other. A rule set catches the obvious ones, a model handles the rest. ### Extract fields **Detail:** Pull text from the body and attachments (OCR scanned files), then extract structured lines with a model, returning a confidence per field. ### Match to MRPeasy **Detail:** Identify the customer or vendor by sender domain and name, then match item codes and descriptions, using customer part number mappings where they exist. ### Create a draft **Detail:** Write a draft customer order, quote or note through the API with the source email id stored, or place it in a review queue if matching is uncertain. ### Review and learn **Detail:** Show the draft beside the source email, let a person correct it, and keep the corrections as examples and as new mapping rules. ### Reply when appropriate **Detail:** Send an acknowledgement from a template only after a human confirms; never auto-reply with prices. ## Example Flows ### PO from a repeat customer **Trigger:** An email with a PDF PO arrives in the orders mailbox. **Steps:** - Sender domain matches a known customer. - Lines are extracted and matched to items via the customer's part number map. - A draft customer order is created with the PO number and dates. - The planner confirms in a review screen. **Result:** Entry time drops to a check, and the PO is traceable to the email. ### Supplier acknowledgement with a new date **Trigger:** A supplier replies that two lines will ship a week late. **Steps:** - Model extracts PO number and new dates. - The system finds the open purchase order. - Proposed date changes are shown to the buyer. - On approval, the date or a note is updated through the API. **Result:** The planner sees the realistic date and MRP reflects it. ## On Prem Emails contain customer pricing and drawings, so many manufacturers do not want them sent to a public model. Run the mailbox reader, OCR and extraction on your own server, and use an open-weight model on local GPUs or a private cloud for classification and extraction. MRPeasy is hosted, so only the structured draft you choose to create travels to its API, while the original email and attachments stay in your own storage. ## Security - Use a dedicated shared mailbox with scoped access, not an executive's mailbox. - Never create confirmed orders automatically; drafts and review for anything with money. - Scan attachments for malware before parsing and ignore macros and executables. - Beware spoofed senders: check domains and, for payment or bank changes, require a second channel. - Retain the source email id with every record for audit. ## Options ### Rules plus manual review **Fit:** Works for a few customers with consistent PO formats. **Effort:** 1 to 2 weeks ### Build with a model in-house **Fit:** Good if you have a developer and varied formats; matching is the hard part. **Effort:** 4 to 8 weeks ### Partner build such as Netray **Fit:** Fits when you want a private model, a review screen and continuing tuning. **Effort:** 4 to 8 weeks ## Pitfalls - Item matching causes most errors; customers use their own part numbers and abbreviations. - Multi-page PDFs and scanned POs lower extraction quality without OCR checks. - Forwarded chains hide the real request in the middle of the thread. - Auto-replying creates loops with other automated senders. - Duplicate POs arrive as reminders and need a duplicate check by PO number and customer. ## Faq ### Can MRPeasy read emails by itself? **Answer:** Not as a general feature to turn inbound emails into orders. It can send documents by email, but reading and understanding incoming mail is done by a separate processor that talks to the MRPeasy API. ### Should orders be created automatically? **Answer:** Start with drafts or a review queue. After some weeks of corrections, you might auto-create for trusted customers with high confidence, but keep a daily review of what the system created. ### What about scanned or photographed POs? **Answer:** They need OCR before extraction and typically lower confidence, so route them to review more often. Clean PDFs from the customer's system are much easier than photos. ### Which email systems are supported? **Answer:** Anything reachable through IMAP, Microsoft Graph or the Gmail API. Microsoft 365 and Google Workspace are the common ones, and shared mailboxes with delegated access are the cleanest setup. ### How long before it saves time? **Answer:** Often a few weeks for the first customer type, mostly spent on item mappings and corrections. Expect the benefit to grow as the mapping table fills, but measure on your own volumes. ## Search terms mrpeasy email automation, mrpeasy purchase order email, email to customer order mrpeasy, automate rfq email mrpeasy, mrpeasy inbox automation, mrpeasy email parsing, mrpeasy supplier confirmation email, mrpeasy api email, mrpeasy order entry automation --- Talk to an engineer: https://netray.co/contact?intent=erp_ai | Machine catalog: https://netray.co/agent/catalog.json