Filters & Exploration

Filters are your tool for drilling down into specific scenarios within your process. Apply one or more filters to narrow your analysis, and watch the process map and all metrics update in real-time.

What Are Filters?

Filters let you narrow down your process analysis to focus on specific cases that match certain criteria. Instead of viewing all 250 loan applications, you can filter to see only "high-value loans that took longer than 14 days" or "rejected applications from Q1 2024."

Why Use Filters?

Without Filters: See the entire process with all cases

• Useful for overall understanding
• Shows general patterns
• Can be overwhelming with diverse cases

With Filters: Focus on specific scenarios

• Compare different cohorts (high-value vs standard)
• Investigate problems (slow cases, rejections)
• Validate hypotheses (do certain attributes affect duration?)
• Share specific views with teammates

Types of Filters

Flow Myna supports several filter types to match different analysis needs:

Duration-Based Filters

Filter by how long cases take:

• "Cases longer than 14 days"
• "Cases between 5 and 10 days"
• "Fastest 25% of cases"
• "Cases exceeding SLA (> 7 days)"

Use cases:

• Identify slow cases for investigation
• Compare fast vs slow processing
• Track SLA compliance

Attribute-Based Filters

Filter by case attributes:

• "Loan amount over $75,000"
• "Risk level = High"
• "Region = Northeast"
• "Customer type = Premium"

Use cases:

• Compare different segments
• Analyze specific product types
• Focus on high-value cases

Event-Based Filters

Filter by presence/absence of events:

• "Has event: Manual Review"
• "Missing event: Credit Check"
• "Has any of: Approval, Conditional Approval"

Use cases:

• Find cases with specific activities
• Identify skipped steps
• Detect process deviations

Path-Based Filters

Filter by the route cases took:

• "Cases following variant A"
• "Cases that went: Application → Credit Check → Approval"
• "Cases with rework loops"

Use cases:

• Analyze specific process paths
• Compare different variants
• Identify exception handling patterns

Time-Based Filters

Filter by when events occurred:

• "Cases started in Q1 2024"
• "Cases completed last month"
• "Events between Jan 1 and Mar 31"

Use cases:

• Period-over-period comparison
• Seasonal analysis
• Track improvements over time

How Filters Work

Filter Application

When you apply a filter:

1. Filter is evaluated against all cases
2. Matching cases are identified
3. Process map updates to show only those cases
4. All metrics recalculate for the filtered set
5. Filter appears in active list (can be toggled or removed)

Example:

• Start: 250 total loan applications
• Apply filter: "Amount > $75,000"
• Result: 62 matching loans (25%)
• Process map now shows only these 62 cases
• Duration statistics recalculate for this subset

Real-Time Updates

All views update instantly:

• Process Map: Nodes and edges show filtered counts
• Variants: Only variants present in filtered cases
• Insights: Recalculated for filtered data
• Statistics: All metrics update

Filter Caching

Flow Myna caches filter results for performance:

• First application: Queries backend (may take 1-2 seconds)
• Subsequent toggles: Instant (uses cached results)
• Navigate away and back: Filter state preserved
• Share URL: Filter included in link

Combining Multiple Filters

The real power comes from combining filters:

AND Logic

Multiple filters combine using AND (all conditions must be true):

Example:

• Filter 1: "Amount > $50,000" → 92 cases
• Filter 2: "Duration > 7 days" → 67 cases
• Combined: Cases that are BOTH high-value AND slow → 31 cases

Layered Analysis

Build up filters iteratively:

Step 1: All cases (250)
↓
Step 2: High-value only (92)
↓
Step 3: High-value + Slow (31)
↓
Step 4: High-value + Slow + Had Manual Review (18)

Each layer narrows the focus, helping you drill down to root causes.

Filter Management

• Active filters shown in filter panel
• Toggle on/off without removing
• Remove filters you no longer need
• Clear all to reset to full dataset

Sharing Filtered Views

One of Flow Myna's most powerful features: filtered views are shareable via URL.

How It Works

When you apply filters:

1. URL automatically updates with filter state
2. Copy the URL from your browser
3. Share with teammates
4. They see the exact same filtered view

Example URL:

/dashboard/project/abc123/map?filters=eyJmaWx0ZX...

The URL encodes:

• All active filters
• Current project context
• Selected cohort (if any)

Use Cases

Collaboration:

• "Hey team, check out these slow high-value cases" [shares link]
• Everyone sees exactly the same filtered data

Documentation:

• Bookmark specific analyses for later
• Include in reports or presentations
• Reference in discussions

Stakeholder Communication:

• Share specific findings without explanation
• Let others explore the same scenario
• Maintain consistent view across team

Learn more: Collaboration

Filter Sources

Filters can be created in multiple ways:

AI Co-Pilot

Ask natural language questions:

• "Show me loans over $75K"
• AI creates and applies appropriate filter

Chart Interactions

Click on charts to filter:

• Click bar in duration histogram
• Select range in timeline
• Pick category in breakdown chart

Process Map

Click nodes or edges:

• Filter to cases through specific node
• Focus on particular transition
• Select variant to highlight

Manual Creation

Build filters explicitly:

• Open filter panel
• Select filter type
• Configure criteria
• Apply to process

Filter Best Practices

Start Broad, Then Narrow

Begin with full dataset to understand overall process, then apply filters to investigate specific scenarios.

Save Important Filters

Bookmark URLs for:

• Regular monitoring views
• Common analysis scenarios
• Stakeholder dashboards

Use Descriptive Names

When sharing, explain what the filter shows:

• ❌ "Check out this link"
• ✅ "Here are the slow high-value loans from Q1"

Compare with Base Case

Always understand:

• How many cases match vs total
• How filtered metrics differ from overall
• Whether patterns are representative

Layer Gradually

Add filters one at a time to understand impact:

• See how each filter reduces case count
• Observe which filters have biggest effect
• Avoid over-filtering to empty set

Next Steps

Master filtering to unlock deep process insights:

Use AI Co-Pilot

• AI Co-Pilot - Create filters through conversation

Explore Process Views

• Process Map - See filtered data visually
• Process Variants - Filter by variant

Advanced Topics

• Collaboration - Share filtered views with team
• Event Groups - Simplify filtered views