Microsoft is working to close the gap between how complex data models work under the hood and how business analysts actually explore them on screen. Tracked under Microsoft's AI at Work roadmap ID 576765 (and corresponding Microsoft 365 Message Center notice MC1484591), Microsoft is extending its customizable Process Intelligence Studio workspace to object-centric process mining (OCPM) within Power Automate Process Mining.
For enterprise IT administrators, Power Platform makers, and process analysts, this update brings a modular, multi-tabbed analytical workspace to multi-object process models—allowing teams to dissect intertwined business entities without flattening their data or jumping between disconnected reporting tools. Here is a complete look at what Process Intelligence Studio brings to OCPM, how the underlying data architecture works, and how to prepare your event logs for deployment.
The Rollout Schedule and Scope
According to Microsoft's AI at Work roadmap (ID 576765) and Microsoft Learn release plan documentation, the integration of Process Intelligence Studio into object-centric process mining is currently listed as In development for Web users in Worldwide (Standard Multi-Tenant) cloud instances.
- Public Preview: Listed on the roadmap and release plan for September 2026.
- General Availability: Targeted for October 2026 on the roadmap, with Microsoft 365 Message Center archive reporting (MC1484591) specifying a general availability target date of October 31, 2026.
- Target Audience & Enablement: Enabled automatically for administrators, makers, marketers, and analysts.
As every seasoned enterprise administrator knows, roadmap schedules represent planning targets rather than ironclad guarantees. Microsoft's release plan documentation explicitly notes that delivery timelines can shift and projected functionality remains subject to change until rollout completes.
Summary: Roadmap item 576765 brings Process Intelligence Studio to OCPM on the Power Automate web portal, targeting general availability by late October 2026 with automatic enablement for admins, makers, marketers, and analysts.
What Process Intelligence Studio Brings to OCPM
Previously, exploring process mining outputs often meant dealing with rigid, fixed overview screens. Microsoft's Process Intelligence Studio (also documented on Microsoft Learn as the process intelligence experience) replaces static dashboards with a customizable, card- and tile-based analytical environment. By extending this workspace to object-centric process mining, Microsoft is giving analysts a single canvas that supports both traditional linear processes and multi-dimensional OCPM models.
According to Microsoft's roadmap specifications and Microsoft Learn documentation, the studio workspace includes several core capabilities:
- Tab-Based Navigation: Analysts can organize distinct analytical angles or process segments into separate tabs (such as Overview, Case Duration, or Variant Analysis) inside a single workspace, switching perspectives without losing context. Each process view—which can be defined in the Process Mining desktop application—can maintain its own independent set of tabs.
- Customizable Grid Layouts: Using the Edit layout control, users can arrange, resize, and position tiles across a flexible grid to compare process maps, charts, tables, and metrics side-by-side.
- A Library of Analytical Tiles: Workspaces can be populated from a component library that includes interactive process maps, performance metrics, attribute breakdowns, case lists, and statistical views.
- Unified Cascading Filters: A dedicated filter bar below the tab bar displays active filters as removable pills. Applying a filter cascades across all tiles and visualizations within the workspace/tab simultaneously, allowing rapid scenario comparisons without reconfiguring individual charts.
- Multi-Dimensional Object-Centric Support: The interface natively renders OCPM models, allowing teams to inspect interconnected business objects, lifecycle relationships, and cross-object metrics on the same canvas.
- Saved Templates and Real-Time Refresh: Teams can save workspace configurations as reusable analysis templates to share standardized views across the organization, complete with a control bar showing the last data refresh timestamp and a manual refresh trigger.
While Microsoft states in its release materials that this fluid workspace reduces training overhead and shortens analysis cycles, the practical takeaway for IT and BI teams is straightforward: it eliminates the need to export data into external tools just to arrange side-by-side scenario comparisons.
Summary: Process Intelligence Studio provides a customizable grid of tiles, multi-tabbed navigation, cascading filters, and reusable workspace templates that now work directly with multi-object OCPM models.
Why Object-Centric Process Mining Matters: OCPM vs. Case-Centric
Why does a workspace update for OCPM matter so much? Because a pretty dashboard is only as honest as the data model underneath it.
In traditional case-centric process mining, every single event in your log must be anchored to one overarching Case ID—for instance, an Order ID. That works fine for a linear IT helpdesk ticket, where one ticket travels from Opened to Assigned to Resolved.
But consider Microsoft's classic retail order-to-cash scenario documented on Microsoft Learn: a customer places two orders (O1, O2), the retailer raises two supplier restock orders (SO1, SO2), issues two invoices (I1, I2), receives a single consolidated payment (P1), and enforces a compliance rule stating that order O2 can only ship after invoices are paid.
| Analytical Challenge | Traditional Case-Centric Mining (Case = Order) | Object-Centric Process Mining (OCPM) |
|---|---|---|
| Data Structure | Every event is forced into a single case notion; one case equals one linear sequence. | A single event can link to multiple objects and object types simultaneously. |
| Gating & Bottleneck Visibility | Requires joining or flattening invoice and payment events into the Order timeline, which can duplicate counts or obscure which invoice gated shipping. | Events link directly to Order, Invoice, and Payment on one map, showing exact convergence bottlenecks. |
| Cross-Department Workflows | Analyzing customer orders alongside supplier restocking requires switching case notions or merging views offline. | Renders both Order and SupplierOrder lifecycles and their touchpoints on the same model. |
| Compliance & Policy Verification | Validating rules that span entities (such as "ship only after payment") requires manual log stitching. | Cross-object dependencies are visible where color-coded object-flow edges intersect. |
When OCPM renders this workflow inside the studio workspace, the process map tracks distinct object lifecycles (with creation and disposal markers per object type), shared activity nodes, and color-coded object-flow edges. You don't have to choose between distorting your metrics or losing half the story. Conversely, if your workflow is tightly scoped with minimal cross-object dependencies, Microsoft notes that traditional case-centric mining remains the simpler, recommended choice.
Summary: OCPM prevents the event duplication and context loss caused by forcing multi-entity workflows into a single Case ID, preserving true relationships across orders, invoices, payments, and shipments.
Under the Hood: Preparing and Ingesting an OCEL Dataset
Process Intelligence Studio is the visualization engine, but getting into it requires feeding Power Automate a properly structured Object-Centric Event Log (OCEL). If your organization wants to test this workspace, your data engineering and Power Platform teams need to structure and ingest the log according to Microsoft Learn's OCPM specifications.
1. Structuring the Target Schema
According to Microsoft Learn guidance on creating an OCEL, the current system supports data ingestion from a UTF-8 CSV file stored in Azure Data Lake Storage Gen2 or Microsoft Fabric OneLake, or a Delta table stored in a Fabric Lakehouse (note that the step-by-step UI ingestion workflow specifically walks through CSV files in Azure Data Lake Gen2 or OneLake).
Your event log must follow the "one row equals one event" rule and include:
- Activity: A concise, human-readable action name (such as
Create OrderorReceive Payment). - Timestamp: Formatted in
yyyy-MM-dd HH:mm:ss(24-hour) or ISO-8601 format, normalized to UTC or a single consistent time zone. - Actor: The user or automated system executing the event.
- Object Type Columns: One dedicated column for each business object type in your scope (for example,
Order,Invoice,Payment,SupplierOrder). Leave columns blank when an event does not touch that object type. - Multi-Object Delimiter: When a single event touches multiple objects of the same type—such as a single payment covering two invoices—separate the IDs with a pipe character (
|), such asI1|I2. - Optional Attributes: Event-level or object-level attributes (such as
PaymentAmount) for deeper financial or operational slicing.
2. Step-by-Step Ingestion and Data Mapping in Power Automate
Once your OCEL CSV is staged in Azure Data Lake Gen2 or OneLake, follow Microsoft's documented ingestion procedure to launch the studio workspace:
- Initiate the Process: Sign in to Power Automate, select Process mining > Start here in the left navigation pane, and select Object-centric process mining.
- Connect Storage: Enter a process name, choose Azure Data Lake or OneLake under Data source, and select Continue.
- Configure Fabric Export (Optional): Choose whether to export the OCPM semantic model to a Microsoft Fabric workspace (requiring you to select a Fabric workspace, a Fabric Lakehouse, and a default report name) or select Skip export.
- Map Required Attributes: After completing the storage connection setup, use the Map your data screen to bind your columns:
- Map one Activity attribute (required).
- Map one Start event timestamp attribute (required).
- Map at least two Object type attributes (required).
- Optionally map End event, Resource, Event level attributes (bound to the event across all referenced objects), and Object level attributes (bound to a specific object type using the value from its first or last referencing event).
- Analyze: Select Save and Analyze. Once ingestion completes, Power Automate navigates directly into the OCPM process intelligence workspace with the Process overview tab active.
Summary: To use Process Intelligence Studio with OCPM, teams must construct an OCEL file where pipe-delimited columns track multiple object types per event, stage it in Azure Data Lake Gen2 or OneLake, and map at least two object types alongside activity and start-event attributes in Power Automate.
Practical Takeaway for Enterprise Teams
Adding Process Intelligence Studio to object-centric process mining is a welcome maturation of Microsoft's Power Automate Process Mining suite. Previously, organizations often faced an awkward divide: case-centric models got the slickest dashboard customization features first, while multi-object OCPM models felt like an advanced science project.
By bringing multi-tabbed navigation, customizable grid layouts, cascading filters, and shareable workspace templates to OCPM, Microsoft is making multi-object analysis practical for everyday business analysts and process owners. However, IT leaders should remember that no workspace layout can fix a messy event log. Before this feature completes its general availability rollout in late October 2026, the smartest move your team can make is auditing your ERP and CRM event pipelines to ensure your cross-object relationships—between orders, shipments, and invoices—are cleanly captured in OneLake or Azure Data Lake.