Analysis
Runs Tab
| Runs List | Lists past and in-progress analysis runs with name, status, date range, and model used. |
| New Run Action | Opens the run creation form to start a new analysis. |
| Run Status Indicator | Shows Pending, Running, Completed, Partially Completed, Failed, or Stopped for each run. |
| Progress Counter | While a run is in progress, shows processed/total conversation counts, polling every 10 seconds. |
| Open Run Results | Selecting a run opens its aggregate charts, insights, and flagged conversations. |
New Run Form
| Analysis Type Selector | Multi-select search dropdown to choose one or more predefined or custom analysis types to apply. |
| Date Range Picker | Sets the inclusive date range of conversations to include, with Yesterday/Last 7/30/90 Days presets. |
| Data Scope (Sampling Mode) | Chooses between All Conversations, Random Sample, Time Stratified, or User Stratified sampling. |
| Sample Size Input | Sets the number of conversations to include when a sampling mode other than All Conversations is selected. |
| AI Model Selector | Chooses the model used to analyze each conversation; each model has a credit-cost ratio. |
| Override Existing Results Toggle | Controls whether cached results are reused (off) or every conversation is re-analyzed from scratch (on). |
| Run Name Input | Optional name for the run; auto-generated (e.g. AR-0001) if left blank. |
| Run Action | Submits the form and starts the analysis, then navigates to the run's results page. |
Run Results — Summary and Insights
| Run Summary Card | Shows run status, date range, conversations covered, and AI model used. |
| Tour / Rerun / Delete Actions | Run-level actions available from the summary card. |
| Overall Insights | Cross-analysis-type conclusion, key patterns, problems, and recommendations; shown when a run has two or more analysis types. |
| Guided Tour | Built-in step-by-step walkthrough of the results page, available on first visit or by clicking the tour button. |
Run Results — Per Analysis Type
| Result Card | One collapsible card per analysis type, expandable/collapsible independently. |
| Analysis Type Status | Shows Pending, Running, Completed, Partially Completed, or Failed for that type within the run. |
| Aggregate Chart | Visual distribution of results across all analyzed conversations for that type. |
| Trend Chart | Shows how results for that type evolved across the date range. |
| Flagged Conversations Action | Opens a drill-down drawer listing conversations the AI marked as warranting attention. |
| Type-level Insights | Conclusion, key patterns, problems, and recommendations specific to that analysis type. |
Drill-Down and Thread Analysis
| Chart/Trend Drill-down | Clicking a chart segment or trend point opens the conversations behind that data point. |
| Thread Analysis Page | Per-conversation, per-analysis-type view showing the AI's overall result, reasoning, and evidence. |
| Message-level Scoring | Each scored message is highlighted with a color and value chip reflecting its contribution to the result. |
| Evidence Panel | Verbatim excerpts the AI cited as support, sorted by relevance, linked back to their source message. |
| Recommendations | Concrete next steps for that conversation; entries marked Configurable can be acted on directly from agent settings. |
Manage Types Tab
| Types List | Grid or list view of all available analysis types, predefined and custom, with filter and sort. |
| Predefined Types | Built-in, read-only analysis types; can be duplicated to create a customized copy. |
| Custom Types | Team-defined analysis types that can be edited, duplicated, deactivated, or deleted. |
| New Type Action | Opens the form to define what to measure and how the result should be expressed. |
| Output Type Selector | Sets the result format for a type: Binary, Score, Classification, or List. |