IFS Cloud and Microsoft Teams integration

What moves between them

The main flow is from IFS Cloud into Microsoft Teams. ml-connector polls IFS Cloud OData endpoints for supplier invoices, purchase orders, and GL journal entries on a configurable schedule (daily, hourly, or tied to finance windows), filters for records in approval or pending states, and posts formatted notification messages to designated Teams channels. Each message includes the key financial data - vendor name, invoice amount, PO number, approval status - and a direct link back to the record in IFS Cloud. No data flows back from Teams into IFS Cloud; Teams acts as an alert and notification surface. The integration respects IFS Cloud rate limits and page sizes and retries failed sends with exponential backoff.

How ml-connector handles it

ml-connector stores IFS Cloud tenant URL, OAuth 2.0 credentials (client_id and client_secret), and Teams Graph tenant ID, client_id, and client_secret encrypted in its database. At each polling interval, it queries IFS Cloud OData endpoints with filters on modified timestamps and entity state fields to identify newly arrived or changed records (e.g., PostingProposalHeadSet with Status=Pending). For each record, ml-connector formats a Teams message containing the document number, vendor or customer, amount, and approval chain status, then POSTs to the Teams channel endpoint via Microsoft Graph. The integration handles OData ETag concurrency requirements by reading records before querying expanded details, respects IFS Cloud's 5000-element page limit by chunking large result sets, and backs off on HTTP 429 rate-limit responses with exponential jitter. Teams webhook subscriptions are renewed before expiry, and failed message sends are logged to the audit trail for replay. IFS Cloud company codes are mapped to Teams channel assignments during setup so each finance center sees its relevant notifications.

A real-world example

A mid-sized manufacturing company runs IFS Cloud for procurement, finance, and field service across three regional plants. The procurement team uses Microsoft Teams daily for cross-functional collaboration. Before the integration, the AP manager had to log into IFS Cloud multiple times a day to check for pending supplier invoice approvals or high-value purchase orders waiting for authorization. With IFS Cloud and Teams connected, new invoices and orders automatically post to the Procurement approval channel with key details and amounts, and the team approves or escalates directly from their existing Teams workflow. Month-end invoice surges that previously required constant vigilance are now visible in real-time, and lost approvals dropped to zero.

What you can do

• •Push new supplier invoices from IFS Cloud to Microsoft Teams as formatted notifications with vendor name, amount, and approval status.
• •Post pending purchase orders to Teams channels so procurement and finance teams stay aligned without logging into IFS Cloud.
• •Poll IFS Cloud OData endpoints on a configurable schedule with automatic retries and full audit logging of every message sent.
• •Map IFS Cloud company codes and cost centers to Teams channels so regional teams see only their relevant financial notifications.
• •Handle OAuth 2.0 credentials, OData ETags, and Teams webhook subscription renewal automatically and securely.

Questions

Which direction does data move between IFS Cloud and Microsoft Teams?

Data flows from IFS Cloud into Microsoft Teams only. ml-connector polls supplier invoices, purchase orders, and GL vouchers from IFS Cloud and posts formatted notifications to Teams channels. No financial data is written back from Teams into IFS Cloud; Teams serves as a notification and approval visibility layer.

Does the integration require IFS Cloud Event Actions or webhooks?

No. Because IFS Cloud Event Actions require manual per-customer configuration and are not API-driven, ml-connector uses pull-based polling of OData APIs with filters on modified timestamps and state fields. This avoids the need for manual IFS admin setup and works with all IFS Cloud instances.

How does ml-connector handle IFS Cloud OData limits and authentication?

ml-connector respects the 5000-element page limit by chunking large result sets and the 1000 requests/minute rate limit by implementing exponential backoff on HTTP 429 responses. It stores the tenant URL and OAuth 2.0 credentials encrypted and refreshes the bearer token before expiry on every polling cycle.