Sage 100 and Monday.com integration

What moves between them

The main flow is Sage 100 into Monday.com. ml-connector polls AP invoices, AP vendors, GL accounts, and purchase orders from Sage 100 on a schedule you define (AP Invoices every 15 minutes, vendors hourly, GL accounts daily, PO every 15 minutes). New or updated records are written to Monday.com as items on designated boards with columns mapped to Sage 100 fields: vendor name, invoice number, amount, due date, PO number, GL account code, and line item detail. GL accounts flow separately so the Monday.com board reflects your complete chart of accounts. There is no return flow: Monday.com is read-only in this integration, so changes made in Monday.com are not written back to Sage 100.

How ml-connector handles it

ml-connector stores Sage 100 SOAP credentials (username, password, company code, IIS server URL) and Monday.com API credentials (personal token or OAuth bearer token) encrypted. On the Sage 100 side, it calls the SOAP endpoint directly for AR and customer data, and for AP, GL, PO, and vendors it routes through the local Windows agent deployed on the Sage 100 server, which wraps the BOI COM interface. Polling uses the DateLastUpdated and DateCreated fields to fetch only changed records, and it tracks the last sync timestamp per board so repeated syncs skip duplicates. Monday.com board and column configuration is read on first sync so the integration maps Sage 100 GL account codes (in their multi-segment format, e.g. 4000-01-00) to the correct board columns, and it respects the board's item structure so each Sage 100 record becomes one Monday.com item with field values set via GraphQL mutation. Because Sage 100 SOAP has no idempotency keys, ml-connector checks for item existence in Monday.com before creating a duplicate, and because Sage 100 has COM record-locking limits, ml-connector implements retry backoff on write failures. Monday.com's rate limits (1,000 to 5,000 requests per minute depending on plan) are tracked and respected; if the board mutation limit (40 per minute) is hit, ml-connector queues updates for the next window. The local Windows agent on the Sage 100 server must be configured with mTLS or API key credentials that ml-connector uses to authenticate; the integration validates the Sage 100 server URL and company code before every call.

A real-world example

A mid-sized distribution company runs Sage 100 on-premises for AP, GL, and purchasing. The finance team uses Monday.com for invoice review and approval workflow, but invoices arrive by email and are typed in manually, and the GL account coding is often wrong or missing because the team can not see the full chart of accounts without jumping back to Sage 100. The integration polls Sage 100 every 15 minutes and creates a Monday.com item for each new AP invoice with vendor name, amount, due date, PO number, and the GL account code from the original PO mapped to a board column. GL accounts are synced daily so the team has a live reference of valid account codes. Invoice review and approval now happen directly on the Monday.com board, and when the team marks an item approved, the finance manager posts the batch to Sage 100 knowing the GL coding is correct because it came from the source record.

What you can do

• •Poll AP invoices from Sage 100 every 15 minutes and create Monday.com items with vendor, amount, due date, and GL account code mapped to board columns.
• •Sync AP vendors and GL account master data daily so your Monday.com board has a live reference of valid accounts and vendor names.
• •Map Sage 100's multi-segment GL account codes to Monday.com columns, handling the customer-specific account structure automatically.
• •Authenticate Sage 100 via SOAP with username and password, route AP and GL queries through the local BOI Windows agent, and authenticate Monday.com with OAuth or personal token.
• •Implement polling intervals tuned for AP invoices, PO, vendors, and GL accounts separately, with deduplication and retry backoff for concurrent write limits.

Questions

How does ml-connector access Sage 100 AP and GL data if SOAP only covers AR?

ml-connector uses the BOI COM layer wrapped by a local Windows agent deployed on the Sage 100 server. The agent bridges the on-premises Sage 100 system and authenticates to ml-connector via mTLS or API key, allowing AP invoices, GL accounts, vendors, and PO records to be queried and written. The SOAP endpoint is used only for AR and customer data.

Why is the Monday.com board read-only in this integration?

Monday.com is a work operating system designed for tracking and collaboration, not as a transaction source system. Syncing from Monday.com back into Sage 100 would require Monday.com to enforce the business rules and multi-segment GL account structure that Sage 100 requires, which would break Monday.com's flexibility. The integration keeps Monday.com as a view and approval layer over Sage 100 data, not a source of truth for GL coding or vendor master records.

What happens when Sage 100 GL accounts use a multi-segment format like 4000-01-00?

ml-connector reads the full multi-segment GL account code from Sage 100 and stores it in a Monday.com column as a text value. When polling new invoices, it preserves the account code from the source PO or invoice line, so every Monday.com item shows the exact GL code without reformatting. The finance team sees the account code and can validate it before approval.