Oracle Fusion Cloud ERP and Tableau integration

What moves between them

The main flow runs from Oracle Fusion Cloud ERP into Tableau. ml-connector polls Oracle Fusion invoices, payments, suppliers, and GL posting records on a configurable schedule, typically every 5 to 15 minutes. Transformed records are written into Tableau datasources, either creating new datasources or appending to existing ones. Tableau can push refresh completion events back to ml-connector via webhooks to confirm that dashboards have updated. Reference data such as GL account lists and supplier master records are synced in the same direction to keep Tableau dashboards working with current chart of accounts and supplier hierarchies.

How ml-connector handles it

ml-connector stores Oracle Fusion OAuth 2.0 Client Credentials encrypted and refreshes the bearer token every 55 minutes to stay ahead of the one-hour JWT expiry. It queries Oracle Fusion using OData filters on LastUpdateDate to detect only new and changed records since the last poll, reducing API load and avoiding duplicate exports. For Tableau, ml-connector exchanges the Personal Access Token for a session token on each datasource publish operation, handling the 240-minute session expiry by refreshing proactively. Since Oracle Fusion Cloud does not support outbound webhooks to external connectors without Oracle Integration Cloud middleware, ml-connector initiates all pulls. When Tableau webhooks fire for datasource refresh completion, they confirm to ml-connector that export workflows succeeded. The connector maps Oracle Fusion invoice and GL account names to Tableau datasource column names, handles pagination across large result sets with OData offset parameters, and stores full audit trails so any record can be re-exported if a Tableau refresh fails partway through.

A real-world example

A mid-sized services firm runs Oracle Fusion Cloud ERP for billing, expenses, and GL, and uses Tableau Cloud for executive dashboards on revenue, margins, and cash flow. Previously, the finance controller extracted invoice and GL reports from Oracle Fusion each month, manually shaped them into CSV files, and uploaded them to Tableau datasources. The dashboards were stale within weeks, and reconciling discrepancies between Oracle Fusion and Tableau required re-tracing exports. With Oracle Fusion Cloud ERP and Tableau connected, every invoice and GL posting change flows to Tableau automatically, dashboards refresh on schedule, and the finance team can drill into current data within minutes of a transaction posting.

What you can do

• •Poll Oracle Fusion Cloud ERP for invoice, payment, supplier, and GL posting changes on a schedule you control, filtering by LastUpdateDate to capture only new and modified records.
• •Create or update Tableau datasources with transformed Oracle Fusion data, mapped to your dashboard's expected column names and data types.
• •Handle OAuth 2.0 bearer token refresh for Oracle Fusion and Personal Access Token session exchange for Tableau, so authentication flows never interrupt exports.
• •Receive Tableau datasource refresh webhooks to confirm successful publish operations, and track state so failed exports can be replayed without duplication.
• •Map Oracle Fusion GL account names, supplier hierarchies, and cost center codes to Tableau dimensions, keeping dashboards aligned with the source of truth in the ERP.

Questions

How does the integration handle Oracle Fusion Cloud ERP's lack of outbound webhooks?

Oracle Fusion Cloud ERP requires polling via REST API rather than webhook push. ml-connector queries the ERP every 5 to 15 minutes using OData filters on LastUpdateDate to detect only new and changed records, reducing API overhead. Tableau webhooks can confirm refresh completion in the opposite direction.

What happens if the Tableau session token expires during a large datasource publish?

ml-connector refreshes the Personal Access Token to obtain a new session token proactively before the default 240-minute expiry window closes. If a publish operation spans multiple calls, the connector obtains a fresh session token for each batch, ensuring no mid-operation authentication failures.

Can ml-connector sync data in both directions between Oracle Fusion Cloud ERP and Tableau?

The integration is designed as one-way: data flows from Oracle Fusion Cloud ERP into Tableau datasources. Tableau is primarily a visualization and analytics platform, not a transactional system, so writes back to Oracle Fusion are not supported. Reference data such as GL accounts and suppliers are read from Oracle Fusion and synced to keep Tableau dashboards consistent.