I think this is a current know issue. From what I can find, the recommendation is Ito choose a lighter coloured them. Other options include, 1) putting the gantt on a canvas with a lighter theme, 2)add a light coloured rectangle or background image behind the timeline Aziz area so the black text has enough contrast

Scolling on iOS works a little differently to windows etc due to Webkit. One thing to look for is a tab of scrolling="no" in the iframe. Even if this is set to no, you can still use the mouse to scroll. Make sure it is set to yes. For iOS you need to set overflow-scrolling: touch. Here's an example <div style="width:100%;…

The documentation hints that the calculation might be different to what you'd expect. It is based on the annualisation factor for hourly rates, and it also interacts with the FTE Annualization Factor (Hourly) = Legal Employer Standard Working Hours per Week × Weeks per Year FTE = Person Working Hours per Week / Legal…

another option is to bypass the semantic layer by using a REST endpoint and use an OAC Action

It depends on the number of records that's being exported. If hit the limits for Excel, etc, one work around is to export to CSV and then import into Excel

See this other forum post, it might be of help

Have you opened an SR for this? If not, consider opening one.

This looks good and informative. Easy to follow and nagivate.

Good idea and would be helpful

How large is the dataset, as this error could occur for large datasets with many columns, and/or with heavy type profiling on save. These can add latency before the confirmation returns

low level alteration to existing functionality, should be easy to implement

Have a look at the browser console to see if there are any errors displayed, try clearing the cache, check for any zeros/negative/null values in the hierarchy, try recreating on a new canvas

Auditing of AI usage is vital

Thanks Ambili and Philippe for this suggestion and feedback. Easier navigation would be useful.

Integration with other LLMs would be interesting to see

This would be useful for consistency

This combination of errors is a strong indicator the job is failing silently before it reads the model. The could be due to a stale or corrupted user session, or a subtle permission gap that isn't captured by role assignment alone. Some things to check/do include, clear the user browser session, check the semantic model…

Here are some links that might be of help Oracle Fusion Agentic Applications announcement (March 2026) Introducing Fusion Agentic Applications – Oracle Fusion Insider blog Oracle AI Agent Marketplace announcement (Oct 2025) Oracle AI for Fusion Applications overview

There is a native connector in OA (https://docs.oracle.com/en/middleware/bi/analytics-desktop/bidvd/data-sources-and-data-types-reference.html#GUID-1DDF52B4-2AE0-4E02-8E9A-EE74D372B1A2) For more real-time CDC you can use Oracle GoldenGate Eleastic Search Handler…

Here are some links to the available pre-built KPIs ERP (Payables, Receivables, GL, etc.): https://docs.oracle.com/en/cloud/saas/analytics/26r1/faiae/payables---ap-prebuilt-kpis.html Full "Using Oracle Fusion Data Intelligence" guide (latest, 26R1): https://docs.oracle.com/en/cloud/saas/analytics/26r1/fawug/index.html…