Conversation Topics provides visibility into why residents, families, and prospects are reaching out. As part of automated post-call analysis, the AI evaluates each completed conversation to classify its primary topic, assess sentiment, and log actionable operational insights.
How Conversation Topics Work in Post-Call Analysis
After an inbound conversation ends, the AI processes the full interaction transcript and audio recording:
Automated Classification: The call is evaluated against your defined topic list and categorized into its primary driver (e.g., Care / Medical Concern, Tour Inquiry, Maintenance).
Multi-Dimensional Analysis: Alongside the topic, the post-call engine extracts:
Call Summary & Actions Taken: A brief overview and logged events (such as live transfers).
Knowledge Gaps: Identifies specific information the assistant or staff could not answer.
Sentiment & Engagement: Rates caller emotion (from -2 Very Negative to +2 Very Positive) and caller engagement level (Low, Medium, High).
Call & Caller Types: Flags caller identity (e.g., Resident, Prospective Resident / Family) and urgency.
Dashboard Metrics & Filtering
The top bar of the Conversation Topics dashboard pairs high-level volume metrics with multi-select dropdown filters, allowing you to isolate specific operational trends across your community.
High-Level Summary Totals
The top-right header displays aggregate counts that dynamically update based on your selected date range and active filters:
Calls: Total number of interactions processed by the system.
Unique Callers: Number of distinct callers, helping identify whether call volume stems from widespread outreach or repeat callers.
Urgent: Total conversations flagged as high-priority, highlighting calls that require immediate attention or escalation.
Date Range Selection
Use the Date Range dropdown to set the evaluation window across all tables and trend charts:
Last 7 Days: Ideal for monitoring recent operational changes or day-to-day call spikes.
Last 30 Days: The default view, best suited for monthly trend reviews and staffing alignment.
Last 90 Days: Designed for quarterly reporting and identifying broader seasonal shifts.
Month to Date: Tracks call volume aligned with your current calendar month.
Segment Filters
Each dropdown filter supports multi-select checkboxes so you can layer criteria and refine your data view:
Topic: Select one or multiple specific conversation topics to isolate department-specific volume, sub-themes, and sentiment.
Channel: Segment conversations by the hardware or communication channel used (e.g., JoyCall, JoyTablet, or JoyTV).
Call Type: Filter by call direction to analyze incoming versus outgoing interactions (Inbound or Outbound).
Caller Type: Narrow analytics based on the caller's relationship to the community (e.g., Resident, Family Member, Prospective Resident, Vendor, or Staff).
Urgency: Filter calls across priority tiers (Low, Medium, High, or Very High) to evaluate how critical situations are handled.
Resolution: Track call outcomes by filtering for completed inquiries, open issues, or interactions flagged for team follow-up (Resolved, Unresolved, or Follow-up Needed).
Viewing & Drilling Into Topic Data
Interactive Sorting: Hover over and click the arrow icon next to any column header to reorder the table:
Topic: Sort alphabetically in ascending (A–Z) or descending (Z–A) order.
Calls: Sort by total volume from highest to lowest or lowest to highest.
% of Total: Sort by category share of total call volume.
vs Prior: Sort by the largest increase or decrease compared to the previous period.
Displays all classified topics ranked in descending order by call volume.
Tracks core operational metrics per category: Calls count, % of Total volume, and period-over-period percentage and numerical shift (vs Prior).
Trends & Sentiment Graphs
Topic Trend Over Time: A daily multi-line chart illustrating call volume trajectories across individual top categories and aggregate other topics.
Sentiment Over Time: Maps daily caller sentiment on a scale from -2.0 (Very Negative) to +2.0 (Very Positive) alongside an aggregate overall sentiment benchmark.
Topic Detail Drawer
Clicking any row in the Ranked Topics table opens a side panel with:
AI Summary: A high-level generated summary showing total calls and the topic's overall share of classified volume.
Common Sub-themes: AI-extracted recurring sub-patterns within that topic.
Calls for this Topic: A chronological log of every conversation tagged with that topic, displaying the Date/Time, Caller Type, and an interaction Summary.
Call Breakdown Modal
Selecting any individual call from the log opens the complete post-call review:
Header Badges: Immediate visual indicators for classified Topic, Knowledge Gaps, call outcome (Successful), and Sentiment.
Tabs & Feedback: Toggle between Overview and Internal notes, or submit staff feedback using the thumbs up/down buttons.
Audio & Transcript: Listen via the built-in media player, download the recording, and review the full timestamped conversation dialogue.
Actions Taken: Real-time status of system actions executed during the interaction (such as connected phone transfers).
AI Diagnostic Breakdown: Detailed evaluations covering:
Success Status: Confirmation of whether the caller's request was fulfilled.
Knowledge Gap: Highlights specific information or schedules the assistant was unable to provide.
Sentiment & Engagement: Qualitative rating of caller tone and engagement level (Low, Medium, or High).
Managing Topics
The Manage Topics modal gives you direct control over the classification rules the AI uses during post-call analysis. Customizing these categories ensures that analytics reflect the specific operational terminology and departments unique to your community.
Accessing Topic Settings
Click the Manage Topics button located in the top-right corner of the dashboard.
Review the list of active categories, including system Default topics and any previously created custom entries.
Creating a New Topic
Click the blue + Add topic button at the top of the modal.
Topic Name: Enter a clear, concise label (e.g., Activities & Outings or Pharmacy Inquiries).
Prompt Description: Provide a short, detailed explanation of what caller requests belong in this category. The AI uses this description directly to evaluate transcripts and accurately classify conversations.
Save your changes to register the new topic into the classification engine.
Editing and Deleting Topics
Edit Topic Rules: Adjust existing descriptions at any time to refine classification accuracy or broaden/narrow what types of calls get tagged under a specific category.
Delete Custom Topics: Click the trash can (Delete) icon next to any custom category to remove it from the classification list.
Note: Updates, additions, or deletions made in Manage Topics apply exclusively to future post-call evaluations and do not retroactively reclassify historical conversation data.