Case Explorer

Overview

The Case Explorer calculator displays case and event attributes in a customizable table format, allowing you to explore process data in detail. This calculator is your primary tool for viewing raw process data, inspecting specific cases, and understanding the exact attribute values that drive your process analytics.

Unlike analytical calculators that aggregate and summarize data, the Case Explorer shows the actual rows of data - either at the case level (one row per case) or at the event level (one row per activity), depending on which columns you select. You control which attributes appear, how the data is sorted, and how many rows to display.

Common Uses

  • Inspect individual cases to understand specific process instances and their attribute values
  • View event-level details to see the complete activity sequence and timestamps for cases
  • Investigate outliers or exceptions identified by other calculators by examining their raw data
  • Validate data quality by reviewing actual attribute values and identifying missing or incorrect data
  • Create focused data views for export to Excel or sharing with stakeholders
  • Build custom reports by selecting specific columns and sorting by key metrics
  • Debug process issues by examining the exact sequence of events and their attributes

Settings

Columns to Display: Select which case attributes and event attributes you want to include in the output table. You can choose any combination of columns from your event log.

  • If you select only case-level attributes (like Case ID, Customer Name, Total Cost), the output shows one row per case
  • If you include any event-level attributes (like Activity Name, Resource, Timestamp), the output automatically shows one row per event
  • Columns appear in the output table in the order you select them
  • Column names must exactly match the attribute names in your event log

Sort Column: Choose which column to use for sorting the results. The sort column must be one of the columns you selected to display. If you don't specify a sort column, the data appears in its natural order from the event log.

Sort Direction: When a sort column is specified, choose the sort order:

  • Ascending: Sorts from lowest to highest (A-Z, 0-9, oldest to newest dates)
  • Descending: Sorts from highest to lowest (Z-A, 9-0, newest to oldest dates)
  • Unsorted: Maintains the original event log order (default if no sort column specified)

Maximum Rows: Specify how many rows to display in the output table. The default is 100 rows to ensure fast performance. Set to a higher value if you need to see more data, or lower if you only want to view the top few results.

  • Values typically range from 10 to 1000 rows
  • The row limit applies after sorting, so you can use this to show "top N" results
  • For example: Sort by duration descending + limit to 20 rows = "20 slowest cases"
  • Lower values improve performance when working with large datasets

Custom Export Settings: (Optional) Configure a custom Excel template for formatted exports. This advanced feature allows you to create branded reports with predefined formatting and layout.

Examples

Example 1: Inspecting High-Value Purchase Orders

Scenario: Your analysis identified purchase orders with unusually high costs. You want to examine the top 25 highest-value orders to understand what makes them different and validate the data.

Settings:

  • Columns to Display: Case ID, Supplier Name, Total Cost, Purchase Category, Approval Date, Status
  • Sort Column: Total Cost
  • Sort Direction: Descending
  • Maximum Rows: 25

Output:

The calculator displays a table with 6 columns and up to 25 rows, showing the purchase orders with the highest costs at the top:

Case ID Supplier Name Total Cost Purchase Category Approval Date Status
PO-2024-8821 Acme Manufacturing $485,200 Equipment 2024-09-15 Completed
PO-2024-9334 TechSystems Global $412,800 IT Infrastructure 2024-10-02 Completed
PO-2024-7892 Industrial Parts Inc $387,500 Manufacturing Supplies 2024-08-28 Completed
... ... ... ... ... ...

Insights: By examining the top 25 cases, you immediately see that high-value orders span multiple categories (Equipment, IT, Manufacturing Supplies) and involve different suppliers. This case-level view helps you understand whether high costs are concentrated with specific suppliers or categories. You can also verify that the data is accurate by checking if the approval dates and statuses make sense for these large purchases.

Example 2: Reviewing Event Sequence for Delayed Cases

Scenario: Several cases are taking much longer than expected. You want to see the complete sequence of events for the 10 slowest cases to understand where time is being lost and which resources are involved.

Settings:

  • Columns to Display: Case ID, Activity Name, Resource, Timestamp, Duration Since Previous Event
  • Sort Column: Timestamp
  • Sort Direction: Ascending
  • Maximum Rows: 200

Output:

Because you included event-level attributes (Activity Name, Resource, Timestamp), the output shows individual events rather than cases. For 10 cases with an average of 15-20 events each, you'll see approximately 150-200 event rows:

Case ID Activity Name Resource Timestamp Duration Since Previous Event
CS-1234 Create Purchase Request John Smith 2024-10-01 08:15 0 hours
CS-1234 Manager Review Sarah Johnson 2024-10-01 14:30 6.25 hours
CS-1234 Procurement Review (empty) 2024-10-08 09:15 162.75 hours
CS-1234 Supplier Selection Mike Chen 2024-10-08 11:30 2.25 hours
CS-5678 Create Purchase Request Alice Wong 2024-10-02 09:00 0 hours
... ... ... ... ...

Insights: The event-level view reveals the complete story of each delayed case. In case CS-1234, you can see that the Procurement Review step took 162.75 hours (nearly 7 days) and had no assigned resource, suggesting this is where the delay occurred. By examining multiple cases, you can identify whether delays consistently happen at the same process step or involve the same resources, pointing you toward systemic bottlenecks.

Example 3: Data Quality Validation for Customer Attributes

Scenario: Before running your main analysis, you want to validate that customer data is complete and correct. You'll review a sample of cases to check for missing values, formatting issues, or inconsistent data.

Settings:

  • Columns to Display: Case ID, Customer ID, Customer Name, Customer Region, Customer Segment, Order Date
  • Sort Column: Customer Name
  • Sort Direction: Ascending
  • Maximum Rows: 100

Output:

A table showing case-level data sorted alphabetically by customer name:

Case ID Customer ID Customer Name Customer Region Customer Segment Order Date
ORD-5634 CUST-0042 (empty) North Enterprise 2024-09-15
ORD-8821 (empty) Acme Industries West SMB 2024-10-02
ORD-3421 CUST-0156 Global Manufacturing Co East Enterprise 2024-08-28
... ... ... ... ... ...

Insights: Sorting by Customer Name immediately reveals data quality issues. Some cases have missing customer names (shown as empty), while others have missing Customer IDs. This type of inspection helps you understand the extent of data completeness problems before running analytical calculators that might produce misleading results due to missing data. You might decide to filter out incomplete cases or investigate the source systems to fix the data extraction.

Example 4: Identifying Rework Patterns in Manufacturing

Scenario: You want to identify cases where the "Quality Check" activity was repeated multiple times, suggesting quality issues or rework. You'll display events for cases sorted by Case ID to see all events for each case grouped together.

Settings:

  • Columns to Display: Case ID, Activity Name, Timestamp, Quality Inspector, Defect Code, Status
  • Sort Column: Case ID
  • Sort Direction: Ascending
  • Maximum Rows: 500

Output:

Event-level table with events grouped by case:

Case ID Activity Name Timestamp Quality Inspector Defect Code Status
MFG-1001 Production Start 2024-10-01 08:00 n/a n/a Started
MFG-1001 Quality Check 2024-10-01 14:30 Jane Lee QDEF-45 Failed
MFG-1001 Rework 2024-10-01 15:00 Bob Smith n/a In Progress
MFG-1001 Quality Check 2024-10-02 09:15 Jane Lee (empty) Passed
MFG-1001 Packaging 2024-10-02 10:30 n/a n/a Completed
MFG-1002 Production Start 2024-10-01 08:15 n/a n/a Started
MFG-1002 Quality Check 2024-10-01 14:45 John Davis (empty) Passed
... ... ... ... ... ...

Insights: By sorting by Case ID, all events for each case appear together in chronological order. You can clearly see that case MFG-1001 had a failed quality check (with defect code QDEF-45), went through rework, and then passed on the second quality check. In contrast, case MFG-1002 passed quality check on the first attempt. This pattern analysis helps identify which defect codes are most common and which inspectors are finding the most issues.

Example 5: Creating Executive Summary of Recent Approvals

Scenario: The CFO wants to see the 50 most recent invoice approvals with key information about amounts, departments, and approvers for a weekly executive summary.

Settings:

  • Columns to Display: Invoice Number, Invoice Date, Department, Invoice Amount, Approver Name, Approval Date
  • Sort Column: Approval Date
  • Sort Direction: Descending
  • Maximum Rows: 50

Output:

Case-level table showing the most recent approvals first:

Invoice Number Invoice Date Department Invoice Amount Approver Name Approval Date
INV-2024-9845 2024-10-18 Marketing $12,450 Sarah Johnson 2024-10-19
INV-2024-9832 2024-10-17 Operations $8,920 Mike Chen 2024-10-18
INV-2024-9801 2024-10-16 IT $45,200 Sarah Johnson 2024-10-18
... ... ... ... ... ...

Insights: This focused view provides exactly what the CFO needs: a chronological list of recent approvals showing who approved what and for which departments. The descending sort by Approval Date ensures the most recent activity appears first. The 50-row limit keeps the report manageable while covering roughly one week of approval activity. This data can be exported to Excel for inclusion in executive reports.

Example 6: Comparing Resource Performance Across Cases

Scenario: You want to analyze how different customer service representatives handle cases by examining their assigned cases and key metrics.

Settings:

  • Columns to Display: Case ID, Assigned Agent, Case Type, Priority, Resolution Time, Customer Satisfaction Score
  • Sort Column: Assigned Agent
  • Sort Direction: Ascending
  • Maximum Rows: 300

Output:

Case-level table grouped by agent:

Case ID Assigned Agent Case Type Priority Resolution Time Customer Satisfaction Score
CS-2234 Alice Johnson Technical Support High 2.5 hours 4.8
CS-2456 Alice Johnson Billing Inquiry Medium 1.2 hours 5.0
CS-2789 Alice Johnson Technical Support High 4.1 hours 4.2
CS-1923 Bob Martinez Account Setup Low 0.8 hours 4.9
CS-2034 Bob Martinez Technical Support High 8.5 hours 3.5
... ... ... ... ... ...

Insights: Sorting by Assigned Agent groups all cases for each agent together, making it easy to see patterns in their work. For Alice Johnson, you can see she handles both technical and billing cases with consistently high satisfaction scores and reasonable resolution times. Bob Martinez shows more variation, with one technical support case taking much longer (8.5 hours) and resulting in a lower satisfaction score (3.5). This case-by-case view helps managers identify coaching opportunities or understand workload distribution.

Output

The Case Explorer calculator displays an interactive data table containing the exact columns you selected, sorted according to your specifications, and limited to the number of rows you defined.

Table Structure

Column Headers: Display the friendly names of the attributes you selected, making the table easy to read and understand.

Data Types: Each column preserves its original data type from the event log (dates, numbers, text, etc.) and formats values appropriately.

Row Count: The table shows up to the Maximum Rows value you specified. If your event log contains fewer matching rows, all available rows are displayed.

Sort Order: Rows appear in the order specified by your Sort Column and Sort Direction settings. If no sorting is specified, rows appear in their natural event log order.

Display Behavior

Case-Level vs. Event-Level:

  • When you select only case attributes, the table shows one row per case
  • When you include any event attributes, the table automatically switches to event-level display, showing one row per event/activity
  • The calculator automatically detects which mode to use based on your column selections

Empty Values: Missing or null attribute values appear as empty cells in the table, helping you identify data quality issues.

Large Numbers: Numeric values format with appropriate thousands separators and decimal places for readability.

Dates and Times: Timestamp columns display in a human-readable format showing both date and time.

Interactive Features

Click-Through: Click on individual rows to drill down into case details and explore the underlying events and attributes in more depth.

Export to Excel: Export the displayed table to Excel for offline analysis, reporting, or sharing with stakeholders who don't have access to mindzieStudio.

Custom Export Templates: If Custom Export Settings are configured, you can generate formatted Excel reports using predefined templates that include branding, formatting, and layout.

Copy Data: Select and copy data from the table for pasting into other applications or documents.

Performance Notes

Row Limits Improve Speed: Using smaller Maximum Rows values (10-100) provides faster results, especially with large event logs containing millions of events.

Event Queries Are Slower: Displaying event-level data (by including event attributes) requires more processing time than case-level data, particularly for event logs with many events per case.

Sorting Performance: Sorting is performed before the row limit is applied, so sorting large datasets may take a few moments. Consider filtering your data with other calculators first if you're working with very large event logs.

Common Output Patterns

Top N Analysis: Sort by a metric column (descending for highest, ascending for lowest) + set Maximum Rows to N = "Top N" or "Bottom N" results

Data Validation: Sort alphabetically by key attributes + review for missing values, duplicates, or formatting inconsistencies

Case Investigation: Select specific Case IDs (using filters first) + include event columns + sort by Timestamp = Complete case timeline

Executive Reporting: Select key business metrics + sort by date descending + limit to recent rows = Weekly/monthly summary for stakeholders

Resource Analysis: Include resource attributes + sort by resource name + include performance metrics = Resource workload and performance review

Usage Tips

Start Simple: Begin by selecting just a few essential columns to understand the data structure, then add more columns as needed.

Combine with Filters: Use filter calculators upstream to narrow down to specific cases of interest before displaying them in the Case Explorer.

Verify Column Names: Column names must exactly match the attribute names in your event log (case-sensitive). Use the column picker in the calculator settings to avoid typos.

Test Row Limits: Start with a small Maximum Rows value (like 10 or 20) to see the table structure quickly, then increase if you need more data.

Event-Level Trade-offs: Remember that including even one event attribute switches the entire table to event-level display, which shows many more rows and may take longer to process.

Export for Sharing: When creating reports for stakeholders, use the Case Explorer to create focused, sorted views, then export to Excel for professional formatting and distribution.

The Case Explorer is your window into the raw process data, providing the transparency and detail needed to validate analytics, investigate specific cases, and understand the exact attribute values that drive your process mining insights.


This documentation is part of the mindzie Studio process mining platform.

An error has occurred. This application may no longer respond until reloaded. Reload ??