Load this month’s actuals from the file I’m uploading.
Finds your saved Data Import job, takes the CSV, TXT or ZIP file, and prepares the import for your approval.
BISP Solutions
Powered By AI
Oracle EPM Planning.
EPM AI Assistant · Oracle EPM Planning
Oracle EPM Planning.
Now with AI.
Ask for the Planning work in plain English.
The assistant prepares it. You approve it.
The EPM AI Assistant finds the matching jobs, rules and pipelines in your live Oracle environment, collects the inputs, and prepares the operation for your approval. It runs through Oracle Planning REST APIs — and every run leaves evidence you can review.
Proposal · for your approval
Metadata Import, then Cube Refresh
- Step 1 · Metadata ImportReady
- Step 2 · Cube RefreshReady
Evidence · every run
Nothing runs without a record
- StatusPolled to completion
- Oracle job IDCaptured
- DiagnosticsReturned
Load this month’s actuals from my file, then refresh the cube.
I found your saved Data Import job and the Cube Refresh job. Here is a two-step flow for your approval.
Run the aggregation rule for the FY27 plan.
I matched the deployed Business Rule and filled in its runtime prompts. Check the values, then approve.
Why did last night’s metadata load fail?
I read that run’s execution history. Oracle’s diagnostics show the step that failed, and why.
What you can ask
Tell it the work. It prepares the operation.
The assistant turns a request into the right Oracle operation — the job, its inputs and the order of steps — and shows you exactly what will run before anything is submitted.
Run the aggregation rule for the FY27 plan.
Matches the deployed Business Rule, asks for its runtime prompts from the RTP registry, and checks every mandatory value.
Which pipelines can I run for Workforce?
Looks through the live Oracle catalog and lists the matching Pipelines, with the runtime variables each one needs.
Why did last night’s metadata load fail?
Reads that run’s execution history and the diagnostics Oracle returned, and points you to the step that failed.
Import the new metadata, then refresh the cube.
Builds a governed two-step flow — Metadata Import, then Cube Refresh — and waits for you to approve it.
Push the forecast to the reporting cube.
Proposes the matching Data Map, with target clearing and member overrides you can adjust before approving.
Example requests. The assistant works with the jobs, rules and pipelines already configured in your own Oracle environment.
How the assistant works
From a sentence to a governed Oracle run
Every request follows the same path, whether it is a single Oracle job or a multi-step flow — and the one step only you can take is the approval.
-
Ask
Describe the work in plain English, the way you would to a colleague.
Plain English -
Understand
The assistant finds the matching jobs, rules, pipelines and integrations in your live environment.
Live environment -
Prepare
It collects the inputs — files, periods, runtime prompts, members — and validates them.
Validated inputs -
You approve
You review the proposal in the workspace. Chat text is never treated as approval.
Only you -
Execute
The operation runs through Oracle Planning REST APIs, on the background worker.
Oracle REST APIs -
Explain
Status, Oracle job IDs, record counts and diagnostics come back to you, kept with the run.
Kept with the run
What it can prepare and run
The Planning operations behind every answer
The assistant works through the platform’s supported operations, using your existing Oracle configurations as the source of truth. Each one can also be run directly from the workspace.
Planning Data Import
Saved Oracle Data Import jobs from CSV, TXT or ZIP files, with the Oracle job ID captured, status polled to completion and parent and child diagnostics returned.
CSV · TXT · ZIP01
Metadata Import
Saved Metadata Import definitions, with an optional Cube Refresh once the load completes.
Optional Cube Refresh02
Data Integrations
File-based Data Integrations through Oracle’s integration REST job, with the definitions and import or export modes already configured in your environment.
Existing definitions03
Oracle Pipelines
Data Integration Pipelines, read from their live definition, with their configured runtime variables prompted for and the process monitored to completion.
Live definition04
Business Rules
Deployed Calculation Manager rules, with runtime prompts served from the RTP registry as ordered, labelled fields and mandatory values checked before submission.
RTP registry prompts05
Data Maps
Plan type map jobs between cubes, including target clearing and member or exclusion overrides, for precise, repeatable pushes to a reporting cube.
Clear · overrides06
Saved Cube Refresh jobs
The Cube Refresh jobs already saved in your Planning application, on their own or as the closing step of a metadata load.
Saved Oracle jobs07
Substitution & User Variables
Oracle substitution variables and Planning user variables, reviewed and updated from one screen, scoped to the connected environment.
Environment-scoped08
Data Explorer & Excel Export
Saved Data Explorer views of live Planning data — inspect, check and compare slices, then export a view as an Excel workbook.
Excel export09
AI with guardrails
It proposes. You approve.
The EPM AI Assistant is permission-aware: it works inside the platform’s own permission checks, so it can only prepare what you are allowed to run. It inspects live Oracle artifact catalogs, answers questions about the available operations, reviews execution history and collects the inputs an operation needs — but work is submitted only after you review and approve the proposal in the workspace.
With your approval, it can
- Inspect live Oracle artifact catalogs and recommend matching artifacts
- Answer questions about available platform operations
- Review execution history and evidence
- Collect required inputs and build governed multi-operation flows
- Run approved operations from the platform’s supported list
- Respect platform permissions on every action it proposes
By design, it never
- Acts on its own — write actions always need your explicit approval
- Creates or changes schedules
- Reaches outside its supported list of operations
- Deletes Oracle artifacts
- Changes Oracle Pipeline definitions
- Generates Data Explorer views from natural language in this release
Permission-aware
Conversation text is never an approval
Every approval kept with its run
Around the assistant
One workspace for the whole Planning team
The assistant sits inside a complete Planning workspace — data checks, people and approvals, scheduling and a full record of every run.
Data Explorer
Pick a cube, define the grid, inspect live values, run data-quality checks, compare source and target slices, save the view and export it to Excel.
Planning Cycles
A Planning calendar with stages, owners, due dates and a My Work queue for each user. Cycles organise people; each Oracle operation is still approved on its own.
Scheduling
Scheduled Oracle Pipeline runs and automatic Business Rule RTP registry synchronization. Everything else runs on demand.
Jobs & Activity
Live status, step-level evidence, Oracle job IDs, record counts where Oracle returns them, downloadable files and logs, and governed retry of supported failed flows.
Access control and governance
Built for auditors as much as administrators
Sign-in
Users sign in with validated Oracle EPM credentials, with local recovery accounts available for platform administration. Oracle user passwords are not retained.
Integration identity
Background operations and schedules use a separately configured Oracle integration identity, stored as a protected server-side secret.
Permissions
Platform permissions govern what users — and the assistant acting for them — may prepare, while Oracle enforces the integration account’s privileges.
Execution history
A lasting record of every execution, approval and diagnostic, kept with the run that produced it.
Ready when you are
Ask for your next Planning run.
Launch the EPM AI Assistant, or talk to the BISP Solutions Oracle EPM team about a guided rollout on your own environment.