Oracle PeopleSoft and Tableau integration

What moves between them

ml-connector polls Oracle PeopleSoft on a schedule you set, reading employee directories, absence records, payroll banking information, and year-end forms. These records are loaded into Tableau datasources via API calls, allowing Tableau workbooks and views to report on current HR and payroll data. The flow is one-way: PeopleSoft to Tableau. Because PeopleSoft is self-hosted and has no webhooks, polling is the only option. Tableau views built on top of these datasources are read-only; ml-connector does not write back into PeopleSoft.

How ml-connector handles it

ml-connector stores both credential sets encrypted: the PeopleSoft hostname, port, node name, and HTTP Basic Auth credentials (or OAuth2 token), and the Tableau Personal Access Token. On each poll, ml-connector connects to PeopleSoft's unique customer-specific REST endpoint, handles date-range filtering to avoid overwhelming the instance, and retrieves employee and payroll records. These are then authenticated to Tableau using the PAT, which ml-connector exchanges for a session token valid for 240 minutes. ml-connector assembles the records into Tableau datasource format and pushes them via Tableau's datasource publish or extract-refresh endpoints. Because PeopleSoft has no rate limits documentation but is capacity-constrained by Tuxedo queue configuration, ml-connector throttles requests and logs each poll so you can monitor performance impact. Credential expiry is tracked so OAuth2 token renewal does not cause outages. Every record carries a full audit trail and can be replayed if a downstream call fails.

A real-world example

A mid-sized financial services firm runs Oracle PeopleSoft for HR, payroll, and GL accounting across multiple departments. Their BI team uses Tableau to build dashboards on headcount, payroll costs by department, and GL account balances for finance leadership. Before the integration, the BI team exported employee and payroll reports from PeopleSoft manually each week, then loaded them into Tableau datasources by hand, which meant dashboards were stale within days and month-end close reporting required last-minute scrambling. With PeopleSoft and Tableau connected, employee and payroll records sync automatically on a nightly schedule, so dashboards are always current and the finance leadership team sees real-time GL balances without waiting for manual exports.

What you can do

• •Load employee directory, payroll, and GL account data from Oracle PeopleSoft into Tableau datasources on a schedule you control.
• •Authenticate PeopleSoft with HTTP Basic Auth or OAuth2 Bearer token, and Tableau with Personal Access Token.
• •Poll PeopleSoft with date-range filtering to minimize impact on the production instance.
• •Track credential expiry and API session state so token renewal does not interrupt the flow.
• •Maintain a full audit trail on every record pulled from PeopleSoft and pushed to Tableau, with replay capability on failure.

Questions

Can ml-connector write HR or payroll data back into PeopleSoft from Tableau?

No. Tableau is a read-mostly analytics platform; its API is designed to manage BI assets and run extracts, not to be a data target for writing back into source systems. The flow is one-way: PeopleSoft to Tableau. If you need to update PeopleSoft records, ml-connector can integrate PeopleSoft with a transactional system instead.

How does ml-connector handle the fact that PeopleSoft is self-hosted with a unique hostname and port for each customer?

ml-connector accepts the full PeopleSoft hostname, port, and Integration Broker node name as customer configuration. Because PeopleSoft publishes no shared base URL, ml-connector stores these per customer and validates endpoints against that specific instance.

Since PeopleSoft has no webhooks, how does ml-connector know when to refresh Tableau?

ml-connector polls PeopleSoft on a schedule you set (daily, hourly, after payroll runs, etc.). Because PeopleSoft is capacity-constrained by Tuxedo queue configuration, ml-connector uses date-range filters to pull only changed records since the last poll, which keeps the integration lightweight and avoids performance impact on your production instance.