AI Receptionist - Independent: Conversations

Every conversation the AI Receptionist has with patients is tracked as a case. Understanding how to monitor cases and navigate Conversation tabs in the console will help you stay efficient with follow-up and maintain patient satisfaction.

Conversation Stages

  • Active Conversations: Cases that are currently in progress or awaiting the next step in the workflow.
  • Actionable Conversations: Cases that require human review or action before they can be completed.
  • Past Conversations: Completed cases that no longer require action but remain available for review.

Knowing how to navigate and monitor cases using Active, Actionable, and Past Conversations can help you stay on top of patient needs and ensure requests are handled efficiently by your agents.

Best Practices

  • Review Actionable Conversations and resolve outstanding requests quickly.
  • Monitor scheduling cases for time-sensitive approval or follow-up.
  • Pay attention to escalation severity to help prioritize workloads.
  • Use conversation transcripts to investigate patient questions or requests.
  • Periodically review Past Conversations to identify workflow improvements.
  • Active Conversations

    Active Conversations is your view of patient cases that are currently in progress. While Actionable Conversations shows cases that need human follow-up, Active Conversations shows every case that is still active in the system. 

    This view provides insight into what your agents are working on and what cases are still in progress. It gives details to help you track, monitor, and verify the AI Receptionist's current tasks.

    Managing Locations

    At the top of the page, you will see a breadcrumb:

    Active Conversations > Location

    This indicates which cases will show in your table view.

    Understanding Filters

    1. Choose a location in the dropdown.

      • Select an account, level, or location
      • All Locations appears for multi-location accounts
      2026-05-20_14-19-14.png
    2. Set a date range for the conversations in the queue.

      • The default view is Yesterday + Today
      • Use presets or custom dates to filter
      • Filters cases based on when they occurred
    3. The case count badge shows how many conversations match your current filters. It automatically updates as filters are adjusted.

  • Actionable Conversations

    Actionable Conversations is your hub for managing AI Receptionist cases that need your attention. It provides a quick view of cases that need human follow-up, like approving appointment requests, contacting patients, or resolving escalations.

    When managing your AI cases, your team can quickly identify what needs action and ensure nothing slips through the cracks by regularly checking Actionable Conversations.

    landing page.png

    Managing Locations & Categories

    At the top of the page, you will see a breadcrumb:

    Actionable Conversations > Location > Category

    This indicates what cases will show in your table view.

    Understanding Filters

    1. Use the Location dropdown to select an account, level, or individual location.

      • Select an account, level, or location
      • All Locations appears for multi-location accounts
      location dropdown.png
    2. Case categories allow you to switch between different types of conversations.

      • Switch between different case types
      • Changing the category resets filters and clears selections
      case category dropdown.png
    3. The date range filter controls which conversations appear in the queue.

      • Defaults to Yesterday + Today
      • Can be adjusted using presets or custom dates
      • Filters cases based on when they occurred
      date range dropdown.png
    4. The case count badge shows how many conversations match your current filters.

      • Displays total cases for the selected view
      • Updates dynamically as filters change
      case count badge.png

    Take Action on Cases

    Actions vary by category. There are three types of actions available: Approve, Deny, and Resolve.

    1. Available in: Scheduling Requests, Rescheduling Requests, and Cancellation Requests

      When a patient submits a request that requires approval, it will appear in the appropriate category for your review.

      How to Approve or Deny a Request

      1. Locate the case in the list
      2. Click anywhere on the case row to open the details panel
      3. Review the request and conversation details
      4. Click Approve or Deny at the bottom of the panel

      You can also take action directly from the list by selecting the checkbox next to a case and choosing Approve or Deny.

      After a request is approved, it is processed and will automatically move forward once confirmed.

      If a request cannot be completed or requires follow-up, it may reappear with a Resolve action.

      approve deny.png

      Key Callouts

      • Approved requests are automatically written back to your system.
      • No additional follow-up is required within Actionable Conversations after approval.
      • Patient confirmations are handled through your existing communication workflows.
    2. Available in: Callback Needed, Denied Appointment Requests, and Missing Information Not Saved

      Once a case has been reviewed and any necessary follow-up has been completed, use Resolve to remove it from your active queue.

      How to Resolve a Case

      1. Locate the case in the list
      2. Click the case to open the details panel
      3. Review the interaction and confirm the issue has been addressed
      4. Click Resolve at the bottom of the panel

      You can also resolve cases directly from the list by selecting the checkbox next to a case and choosing Resolve.

      resolve case.png
  • Past Conversations

    Past Conversations provides a history of completed interactions handled by the AI Receptionist. When a case is resolved, it is automatically moved here for future reference.

    Managing Locations

    At the top of the page, you will see a breadcrumb:

    Past Conversations > Location

    This indicates what cases will show in your table view.

    Understanding Filters

    1. Date Range

      • Adjust using presets or custom dates
      • Filters conversations based on when they occurred

      A count badge next to the date range shows the total number of conversations matching your current filters.

      combined location and date range filtering.png
    2. Reasons Filter

      • Filter conversations by escalation type (e.g., Emergency, Follow-Up Requested)
      • Each option shows a count of matching conversations
      • Select again to clear the filter
      filter by reasons.png

      Tags Filter

      • Filter by outcome tags (e.g., Scheduling Succeeded, Cancellation Rejected)
      • Search to quickly find specific tags
      • Tags are color-coded by sentiment
      filter by tags.png
    3. Tags are automatically applied to each conversation to show the outcome. They help you quickly understand what happened without opening the case.

      Common Tags

      • Scheduling Succeeded — Appointment booked
      • Scheduling Rejected — Request denied
      • Scheduling Without Approval — Booked without review
      • Rescheduling Succeeded — Appointment updated
      • Cancellation Succeeded — Appointment cancelled

      Tag Colors

      • Green — Positive outcome
      • Red — Negative outcome
      • Grey — Neutral outcome
    4. Past Conversations only includes completed cases. No action is required when reviewing these conversations.

      Each case includes a status to indicate how it was resolved:

      • Completed — The request was successfully handled
      completed case.png
      • Closed — The request was not completed
      closed case.png
      • Cancelled — The interaction was stopped by the system or staff
  • Case Lifecycle

    Cases progress through different states (Active, Actionable, Past) as agents carry out conversations, answer questions, and solve patient requests. Follow the lifecycle to understand what case progression looks like.

    Conversation Begins

    When the AI agent starts a conversation with a patient, a case is created. This happens when a patient:

    • Requests to schedule, reschedule, or cancel
    • Has missing information in their profile
    • Asks questions that require follow-up

    Once created, the conversation becomes visible in Active Conversations.

    Conversation Progresses

    As the AI Receptionist communicates with a patient, cases will move to Actionable Conversations if human follow-up is needed. Here, users can take action to keep the case moving along (Approve, Deny, or Resolve). 

    Cases that need intervention are still visible in Active Conversations, but action is available only from Actionable Conversations. Cases will remain in Active Conversations until they reach termination.

    Conversations Ends

    A case moves from Active Conversations to Past Conversations when:

    • The case goal has been achieved
    • The case is cancelled by the staff or the system

    Case Details

    Selecting a case opens the conversation details, where you can review the full interaction and any related patient or request information.

    Conversation Transcripts

    Click View in the Transcript column from the case list:

    • Opens the conversation script without opening the case panel
    • Includes timestamps and message attributions

    convo transcript.png

    This view gives your team the context needed to understand what happened and determine the appropriate next step.

    Clicking a case from the list opens a side panel with three tabs.

    1. The Overview tab provides a summary of the case, including key patient details and request information.

      • Quality Tags — Tags related to conversation quality
      • Summary — A brief description of the last interaction, generated by the AI agent 
      • Evaluation Rating — A quality rating (1-5 dots) for the interaction 
      • Escalations — A list of active escalations, each showing: 
        • Escalation type (e.g., “No Engagement”, “Emergency”) 
        • Severity badge (Critical or Moderate) 
        • Status (Active or Resolved) 
        • Description — click to expand for full details 
      • AI-Extracted Data — Expandable sections showing data the AI agent gathered during the conversation: 
        • Scheduling Details — Appointment type, date/time, provider, and whether booking succeeded or failed 
        • Rescheduling Details — Previous and new appointment details 
        • Cancellation Details — The appointment to be cancelled 
        • Patient Details — Verified patient information (name, date of birth, phone, address, insurance) 
        • Appointment Preference — What the patient requested (type, time, provider, location) 
        • Benefits — Insurance and pharmacy information 
        • Transfer Details — If the call was transferred, the target number and status 
        • Missing Information — PMS writeback status (Pending or Failed) when applicable 

      This is the best place to quickly understand the context of the case before taking action.

      Overview tab.png

      Quality Tab Colors

      Tags help customers quickly identify what happened in an interaction:

      • Green (Positive) = Scheduling succeeded, Rescheduling succeeded, Confirmed
      • Red (Negative) = Scheduling rejected, Cancellation rejected, No response
      • Grey (Neutral) = Scheduling without approval, Patient verified
    2. The Interactions tab displays the full conversation between the patient and the AI Receptionist:

      • A chronological timeline of all interactions (calls, messages) 
      • Each interaction shows the channel (Voice, SMS, Chat), timestamp, and duration 
      • Click an interaction to view the full transcript with individual messages 

      Interactions tab.png

       

    3. The Reference tab provides background information:

      • Patient Summary— Full patient demographics (name, date of birth, phone, address, insurance details) 
      • Case Metadata — Case creation date, last update, external IDs, clinic name, and other system information 

      Reference Tab.png

       

  • Navigation

    Here's your guide to navigating the elements of the Conversation tabs in the console.

    Categories

    Conversations are grouped into categories to help your team quickly identify which cases need attention.

    CategoryWhat It ContainsAction
    Callback NeededCases where patients requested a callback because the AI agent couldn't resolve the requestResolve
    Scheduling RequestsCases where patients requested an appointment that requires staff approval to bookApprove, Deny, Resolve
    Rescheduling RequestsCases where patients requested to reschedule an existing appointment, which requires approvalApprove, Deny, Resolve
    Cancellation RequestsCases where patients requested to cancel an upcoming appointment, which requires approvalApprove, Deny, Resolve
    Denied Appointment RequestsCases where patient appointment requests were deniedResolve
    Direct Scheduled AppointmentsCases where patient appointments were directly scheduledNone (review only)
    Missing Information Not SavedCases where the collection of missing patient information was unsuccessfulResolve

    Actionable Conversations include a checkbox with the option to approve, deny, or resolve selected cases. 

    Columns

    Columns provide key details at a glance, such as the patient, case type, request stage, assigned user, and most recent activity. Each row represents a patient case. The following columns are displayed:

    • Name — Patient full name
    • Phone — Patient phone number
    • Location — Associated location for this case
    • Type — Patient type (New or Existing)
    • Apt Type — Appointment type
    • Apt Date/Time — Scheduled appointment date and time
    • Modality — Communication channel used for this interaction
    • Interaction — Timestamp of the last interaction
    • Transcript — View the full conversation
    • Tags — Outcome tags showing the interaction's result
    • Escalation — Severity level for escalated cases (Critical or Moderate)

    columns.png

    The columns icon on the left side of the case list allows you to edit the column view. 

    • Select a column name to hide or show it
    • Checkmarks indicate visible columns
    • No checkmarks indicate hidden columns

    edit columns view.png

    Navigation Tips

    • Column Customization: Use the column icon to show or hide certain fields. Selections will remain until the category is changed or the page is refreshed.
    • Row Eligibility: Not all rows are selectable. A checkbox appears only if the case is in an active state.

    Search

    Type in the search box at the top of the list to find cases based on status, category, or other details. 

    Searches:

    • Patient name
    • Phone number
    • Case-related fields
    search option.png

     

    Escalation Reason Filter (Callback Needed Only)

    The Callback Needed category has its own unique Reasons filter. This allows you to filter cases by escalation reason:

    • No Engagement 
    • Emergency
    • Goal Not Met
    • User Concerns
    • Follow-Up Requested
    • Wrong Recipient

    Escalation filter dropdown.png

    Reset Filters

    A reset option appears in the top right corner when filters are active. Resets will clear all chosen filters except for Location and Category choices.

    reselt filters.png

    Refresh Button

    Located in the top right corner of the list, this button refreshes the current data view.

    refresh button.png

    Pages

    The table shows 25 cases per page. Use the Previous and Next buttons at the bottom to navigate between pages. The current page and the total number of pages are displayed between the buttons. 

    • Switching categories clears filters and selections
    • Changing pages clears selected rows
    • Changing location preserves agent selection