22.1 Overview
AI Meeting is AIUNIFY PROS SUITE’s real-time video-meeting workspace for creating and joining browser-based meetings with video, audio, participant Chat, screen sharing, live transcription, and an optional spoken AI Agent.
The Lobby identifies the workspace as:
AI Meeting
with:
“Create or join secure video meetings instantly. Crystal-clear audio and video powered by WebRTC peer-to-peer technology.”
The current source verifies subscriber-facing functionality for:
AI Meeting is fundamentally a live communications workspace, not the AI Meeting Notes Generator.
22.2 Important Scope Distinction
Three different AIUNIFY capabilities should remain separate:
AI Meeting
The PROS SUITE live Meeting workspace documented in this chapter.
Meeting Notes
A separate Business Tool that accepts a transcript or notes and generates:
AIUNIFY Video Conferencing
AIUNIFY’s separate Video system is outside this PROS SUITE chapter.
AI Meeting should therefore not be described as automatically generating the structured Meeting summaries handled by the separate Meeting Notes Tool.
22.3 Where AI Meeting Is Located
Navigate to:
Sidebar → Core Tools → AI Meeting
The current route is:
/ai-meeting
and the Sidebar marks it as a New Core Tool.
22.4 Authentication Is Required
Both:
/ai-meeting
and individual Meeting rooms at:
/ai-meeting/[meeting-id]
require an authenticated PROS SUITE session.
If you are not signed in, the current route sends you toward:
/login.
22.5 Meeting Links Are Not Verified as Anonymous Guest Links
This is an important current-version rule.
A person opening a Meeting link is still routed through authentication and:
FeatureGuard → ai-meeting.
Therefore:
The current source does not establish anonymous guest Meeting participation.
Invitees should be prepared to sign into an account with access to AI Meeting.
22.6 Tool Access Still Applies
AI Meeting's Lobby and Meeting Room are protected by the normal PROS SUITE feature-access system.
Therefore, having a valid Meeting link does not itself bypass:
22.7 AI Meeting Lobby
The AI Meeting Lobby contains:
This is the central starting point for Meeting activity.
22.8 Real-time Video & Chat
The Lobby displays:
Real-time Video & Chat.
This accurately describes the two primary communication methods available inside the Meeting Room.
22.9 Current Feature Highlights
The Lobby currently advertises:
22.10 “Up to 8” Is the Current UI Capacity Label
The visible Lobby says:
Up to 8 Participants.
The included client source does not independently show a hard participant-count rejection at eight users.
Accordingly, this manual records:
Up to 8
as the current user-facing capacity designation.
22.11 Encryption Label
The Lobby also displays:
Encrypted — End-to-end.
The source does verify a WebRTC peer-connection architecture, but this User Operations bundle is not a complete independent security audit.
Therefore, the manual records the current interface designation without expanding it into unsupported infrastructure guarantees.
22.12 Creating a New Meeting
The first Lobby Card is:
New Meeting
with:
“Start a meeting instantly”.
22.13 Meeting Title
The New Meeting Card provides:
Meeting Title (optional).
Example placeholder:
e.g. Weekly standup
22.14 Meeting Title Is Not Required
The creation request sends a Title only when one was entered.
You can therefore create an instant Meeting without naming it first.
For organization and Recent Meeting identification, however, using a meaningful Title is recommended.
22.15 Useful Meeting Titles
Examples:
A descriptive Title makes Recent Meetings easier to identify.
22.16 Create & Join
When ready, select:
Create & Join.
The Meeting is created and, on success, PROS SUITE routes you to the generated Meeting Room.
22.17 Creating State
While Meeting creation is underway, the button changes to:
Creating...
22.18 Meeting Creation Creates a Join Link
The Meeting creation service returns a:
joinLink
and the Lobby navigates to it automatically after successful creation.
You do not need to manually construct the Meeting ID.
22.19 Meeting Creation Is Immediate
The current Lobby does not contain fields for:
AI Meeting's verified creation workflow is:
Create an instant Meeting now.
22.20 No Scheduling Workflow Is Established
There is no source-verified:
Schedule Meeting
control in the current AI Meeting Lobby.
If a Meeting needs calendar scheduling, use the appropriate scheduling/calendar workflow separately.
22.21 Joining an Existing Meeting
The second Lobby Card is:
Join Meeting
with:
“Enter a meeting ID or link”.
22.22 Meeting ID or Link
The Join field is labeled:
Meeting ID or Link.
The example placeholder is:
e.g. abcd-efgh-ijkl
22.23 You Can Paste the Full Meeting Link
The current Join workflow is designed to extract the Meeting ID from links containing:
/ai-meeting/[id].
This means an invitee does not need to manually isolate the ID from a copied Meeting link.
22.24 You Can Also Enter Only the Meeting ID
If you already have just the ID, enter it directly.
The Join workflow accepts alphanumeric characters and hyphens.
22.25 Enter Can Join the Meeting
While the Meeting ID/Link field is active, pressing:
Enter
invokes the Join Meeting workflow.
22.26 Join Meeting Button
You can also select:
Join Meeting.
The button remains disabled while the input is empty.
22.27 Meeting ID Validation
The current client requires the resulting Meeting ID to:
22.28 Invalid Meeting ID
If validation fails, the exact current message is:
“Invalid meeting ID or link. Please check and try again.”
22.29 Query Strings and Anchors Are Removed
If a pasted Meeting link contains:
the Join workflow cleans those portions from the extracted Meeting ID before validating it.
22.30 Recent Meetings
The lower part of the Lobby is titled:
Recent Meetings.
The Lobby retrieves Meeting records from the Meeting service when the page loads.
22.31 Recent Meetings Are Persistent Records
Unlike in-room Chat and live Transcription, the Meeting Lobby does have a source-verified Meeting list loaded from:
/api/meetings.
A Meeting record contains:
22.32 No Meetings Yet
If the Meeting list is empty, the Lobby displays:
No meetings yet
and:
“Create your first meeting to get started”.
22.33 Meeting Record Information
A Recent Meeting row displays:
22.34 Live Meetings
An active Meeting displays:
Live
with an animated indicator.
22.35 Active Meeting Rows Can Be Reopened
Selecting a Recent Meeting row opens the room only when its status is:
active.
22.36 Non-Active Meeting Records Are Not Reopened from the Row
The current row-click behavior does not navigate into a Meeting unless its status is active.
This prevents the Recent Meetings list from acting as a playback/archive system for completed Meetings.
22.37 Copy Meeting Link from Recent Meetings
Hovering a Recent Meeting exposes a Copy control.
Selecting it copies:
[current origin]/ai-meeting/[meeting-id]
to the clipboard.
22.38 Copy Confirmation
After copying, the Copy icon temporarily changes to a Check indicator for approximately two seconds.
22.39 Opening a Meeting Room
Once you create or join a Meeting, AI Meeting first retrieves the Meeting's information.
The Room can move through:
states.
22.40 Loading Meeting
While Meeting information is being retrieved, the interface displays:
Loading meeting...
22.41 Meeting Not Found
If the Meeting service cannot retrieve the requested Meeting, the Room enters its Error state.
One possible underlying error is:
Meeting not found.
22.42 Ended Meeting
If the returned Meeting status is:
ended
the Room rejects entry with:
This meeting has ended.
22.43 Unable to Join
The Error screen is titled:
Unable to Join.
It displays the associated Meeting Error below the heading.
22.44 Back to Lobby
From an Unable to Join screen, select:
Back to Lobby.
22.45 Retry
The Error screen also provides:
Retry.
This reloads the current Meeting page.
22.46 Pre-Join Screen
Before connecting to the Meeting, the Room displays a Pre-Join screen containing:
22.47 Camera and Microphone Permission Is Requested Automatically
When the Room reaches Pre-Join, AI Meeting automatically attempts to request browser permission for:
and then create a local media Preview.
22.48 Browser Permission Prompt
Your browser may ask whether AIUNIFY PROS SUITE can access:
Allow access when you intend to use video/audio in the Meeting.
22.49 HTTPS/Secure Context Requirement
If browser media APIs are unavailable, AI Meeting can display:
“Camera and microphone access is not available in this browser. Please ensure you are using HTTPS or localhost.”
22.50 Permission Denied
If browser access was denied, the exact current Error is:
“Camera and microphone access was denied. Please allow access in your browser settings.”
22.51 No Camera or Microphone Found
If no suitable hardware is detected, AI Meeting can report:
“No camera or microphone found. Please connect a device and try again.”
22.52 General Media-Device Error
Another current Error is:
“Failed to access media devices. Please check your browser settings or ensure secure context (HTTPS).”
22.53 Local Camera Target
The current media setup requests an ideal Camera configuration of approximately:
These are requested ideals, not guaranteed output specifications for every device/network.
22.54 Audio Processing Requests
The current microphone constraints request browser media processing including:
Actual behavior remains dependent on browser/device support.
22.55 Camera Preview
When camera access is available and Video is enabled, the Pre-Join screen shows your local video Preview.
The local Preview is horizontally mirrored.
22.56 Camera-Off Preview
When Video is unavailable or turned off, AI Meeting displays an avatar-style circle containing the first character of your user Name.
22.57 Pre-Join Microphone Control
Use the Microphone button to choose whether you enter the Meeting:
22.58 Pre-Join Camera Control
Use the Camera button to choose whether Video is:
before joining.
22.59 Joining as
Below the Join button, AI Meeting displays:
Joining as [Name].
22.60 How Your Meeting Name Is Chosen
The current Meeting Room derives your display Name from:
Because the room requires authentication, normal subscribers should generally see their account Name or Email-derived Name.
22.61 No Name-Edit Field Is Established on Pre-Join
The current Pre-Join source does not provide a:
Display Name
editing field.
Your Meeting Name is derived from your current PROS SUITE account information.
22.62 Join Meeting
When ready, select:
Join Meeting.
22.63 Joining State
While the transition is occurring, the button displays:
Joining....
22.64 Meeting Connection Architecture
The current Room uses browser:
WebRTC
peer connections for audio/video streams.
WebRTC connections are coordinated using the Meeting's signaling system.
22.65 Peer-to-Peer Media
Each participant's browser creates WebRTC peer connections to other Meeting participants and adds the local media tracks to those connections.
This aligns with the Lobby's:
Peer-to-Peer
designation.
22.66 Connected Meeting View
After joining, the interface becomes a full-screen-style Meeting workspace with:
22.67 Meeting Status Indicator
The top-left Meeting header shows an animated green status dot next to the Meeting Title.
22.68 Meeting ID Remains Visible
The current Meeting ID also appears in the top bar while connected.
22.69 Invite
The top bar provides:
Invite.
Selecting it copies the current Meeting link.
22.70 Copied!
After the link is copied, Invite temporarily changes to:
Copied!
22.71 Sharing the Meeting Link
A typical workflow is:
22.72 Participant Video Grid
When no screen is being shared, AI Meeting arranges:
in a responsive Meeting grid.
22.73 Grid Adapts to Participant Count
The source adjusts the Grid depending on participant count:
22.74 Participant Count Includes AI Agent
The displayed participant count is calculated as:
Remote peers + You + AI Agent when active.
Therefore, activating the AI Agent can increase the visible Participant count by one.
22.75 Local Participant Tile
Your tile displays:
22.76 Remote Participant Tiles
Remote participant tiles can reflect:
22.77 Speaking Indicator
Participant tiles analyze local audio levels and can visually highlight a tile when audio activity suggests the participant is speaking.
This is a visual aid rather than a speech-transcription feature.
22.78 Meeting Controls
The bottom Meeting toolbar provides core controls for:
22.79 Microphone — Mute
When your Microphone is active, the tooltip reads:
Mute.
22.80 Microphone — Unmute
When muted, the control changes to:
Unmute.
22.81 Muting Changes the Audio Track
The Meeting Room disables/enables your outgoing local Audio track and broadcasts the updated media state to connected participants.
22.82 Camera — Turn Off
When Video is active, the Camera tooltip says:
Turn off camera.
22.83 Camera — Turn On
When Video is disabled, the control says:
Turn on camera.
22.84 Remote Participants See Your Media State
Microphone, Camera, and Screen Sharing states are broadcast to other connected peers so their interface can reflect your current status.
22.85 Screen Share
On desktop, select:
Share screen.
Your browser then controls which available Screen, Window, or Tab can be shared, according to its own sharing dialog.
22.86 Browser Screen-Sharing Permission
AI Meeting uses the browser's screen-capture function.
The browser—not PROS SUITE—determines the exact choices displayed in the screen-sharing selection dialog.
22.87 Screen Sharing Does Not Request System Audio
The current screen-share request specifies:
Therefore:
The current source does not establish sharing computer/system audio through Screen Share.
22.88 Your Microphone Remains Part of the Shared Stream
When screen sharing begins, AI Meeting constructs a new outgoing stream containing:
22.89 Shared Screen Becomes the Main View
When a participant is sharing a Screen, the Room changes layout.
The shared Screen becomes the primary large area while the remaining participants appear as smaller thumbnails alongside it.
22.90 One Active Screen Share at a Time
If you are screen sharing and another participant begins a Screen Share, the current Room automatically stops your local sharing operation.
This supports one active shared-screen focus at a time.
22.91 Stop Sharing
When sharing is active, the control changes to:
Stop sharing.
22.92 Stopping Through the Browser
If you select the browser's own:
Stop sharing
control, AI Meeting detects the end of the screen track and restores your original camera stream.
22.93 Participants Panel
Select:
Participants.
The panel opens on the right side of the Room.
22.94 Participant Count
The header displays:
Participants ([count]).
22.95 Your Participant Entry
Your row displays:
[Your Name] (You)
along with your Microphone status.
22.96 Remote Participant Entries
Each remote participant entry displays:
22.97 No Remote Mute Control Is Established
The Participants panel displays status but does not provide the local user with a verified control to:
22.98 No Host-Management Panel Is Established
The current source does not define participant roles such as:
inside the user interface.
All connected participant tiles are handled primarily as peers.
22.99 Meeting Chat
Select:
Chat.
This opens the:
Meeting Chat
panel.
22.100 Chat and Other Side Panels Are Mutually Exclusive
Opening Meeting Chat closes:
Likewise, opening Participants or Transcription closes the other side panels.
22.101 Empty Chat
Before messages exist, Meeting Chat displays:
No messages yet
and:
Start the conversation!
22.102 Chat Input
The Meeting Chat field displays:
Type a message...
22.103 Sending Chat Messages
You can:
to send a Meeting Chat message.
22.104 Shift + Enter Behavior
The current Chat handler only sends on Enter when Shift is not held.
However, the input itself is a single-line Input rather than a multiline Textarea.
Do not rely on Meeting Chat for composing long multi-paragraph messages.
22.105 Chat Message Information
Messages display:
22.106 Chat Uses Peer Data Channels
Meeting Chat messages are transmitted directly through open WebRTC Data Channels to connected peers.
The sender also adds their own message into their local Chat display.
22.107 Meeting Chat Is Not Source-Verified as Persisted
Chat messages exist in the current Meeting Room's client state and peer Data Channels.
The included source does not establish database persistence of Meeting Chat.
22.108 No Saved Chat History
The current AI Meeting source does not establish:
Treat Meeting Chat as a live-session communication channel.
22.109 No File Attachments in Meeting Chat
The current Meeting Chat supports Text.
It does not establish:
inside the Chat panel.
Use Screen Share or another appropriate system for visual/document collaboration.
22.110 Real-Time Transcription
AI Meeting includes:
Transcription.
Selecting it opens the:
Live Transcript
panel.
22.111 Live Transcript
The panel header is:
Live Transcript.
22.112 Waiting for Speech
Before transcript entries exist, the panel displays:
Waiting for speech...
22.113 Transcript Entry Information
Each transcript entry contains:
22.114 Transcript Automatically Scrolls
As new speech entries arrive, the Live Transcript panel automatically scrolls toward the latest entry.
22.115 Transcripts Are Not Stored Permanently
The interface explicitly states:
“Transcripts are generated in real-time and not stored permanently.”
This is a critical current-version limitation.
22.116 No Transcript Export Control
The Live Transcript panel does not establish:
Treat the current Transcript as temporary live-session information.
22.117 Transcription Uses Browser Speech Recognition
The current implementation uses:
from the browser.
This means Transcription support can vary by browser.
22.118 Current Transcription Language
The speech-recognition engine is currently configured to:
en-US.
The current user interface does not expose a transcription-language selector.
22.119 Continuous Recognition
Speech recognition is configured for:
when the transcription/Agent engine is active.
22.120 Natural Pause Detection
If speech recognition has not yet returned a final phrase, the current implementation waits approximately:
1.8 seconds of silence
before processing the accumulated speech.
22.121 Transcription Is Distributed to Other Participants
When a final local transcript is generated, AI Meeting sends it to connected peers through the Meeting's WebRTC Data Channels.
22.122 Each Participant Generates Their Own Local Speech Recognition
The architecture does not transcribe a centralized mixed Meeting recording.
Each active client recognizes speech through its own browser and broadcasts recognized transcript entries.
This is consistent with the peer-based architecture.
22.123 Transcription Availability Depends on Browser Support
If the browser does not provide the Speech Recognition API, the current Agent/transcription engine records an Error state.
Use a compatible current browser when Live Transcript is important.
22.124 AI Agent Mode
AI Meeting includes an optional:
Agent Mode.
This adds an AI Agent participant tile to your current Meeting interface.
22.125 Turning Agent Mode On
Select the Bot control labeled:
Agent Mode.
When active, the tooltip changes to:
Turn off Agent.
22.126 AI Agent Appears as a Participant
When Agent Mode is active, AI Meeting adds an:
AI Agent
tile to the participant grid.
22.127 AI Agent Statuses
The AI Agent supports the visible states:
22.128 Listening
When listening, the AI Agent tile displays the current recognized speech transcript when available.
22.129 Processing
When the AI is generating its response, the tile changes to:
processing
with a loading indicator.
22.130 Speaking
When the generated response is being read aloud, the tile changes to:
speaking.
22.131 AI Agent Uses the Shared PROS SUITE AI Model
Unlike the core WebRTC Meeting itself, the AI Agent reads your current shared:
selectedModel
from the PROS SUITE authentication context.
Therefore, Chapters 10 and 11 can affect the AI Agent portion of AI Meeting.
22.132 AI Agent Uses User Provider API Keys Where Applicable
The AI Agent adds configured provider API Keys to its AI-stream request.
If your selected Model requires a provider credential, configure it through the normal PROS SUITE Settings workflow.
22.133 AI Agent Default Fallback
If no Model value is available to the current Agent request, its client request falls back to:
gemini-2.5-flash.
This Agent-specific fallback should not be confused with the general model catalogue's normal current default behavior documented in Chapter 10.
22.134 AI Agent Maintains Conversational Context During Its Active Session
The local Agent maintains a message history of:
and includes that history in its next AI request.
Therefore, unlike Live Chat in Chapter 21, AI Meeting's Agent has source-verified in-session conversational context.
22.135 Agent Conversation Is Cleared When Agent Mode Starts
When Agent Mode is toggled on and the Agent session begins, its internal message history is cleared.
A newly activated Agent therefore begins a fresh conversational context.
22.136 AI Agent Is Designed for Concise Spoken Answers
Its system instructions tell the Model to:
because the response is spoken immediately through text-to-speech.
22.137 AI Agent Voice Uses the Browser
The Agent uses the browser's:
SpeechSynthesis
system for spoken replies.
It does not source-verify a separate cloud TTS service for this Meeting Agent workflow.
22.138 Voice Selection Is Automatic
The current Agent attempts to select an English browser voice using a preference sequence including:
22.139 No Manual AI Agent Voice Selector
The current Meeting interface does not expose a user control to choose:
for the Agent.
The source currently uses:
22.140 AI Agent Listening Resumes After Speaking
After the Agent finishes speaking, it automatically resumes its listening state while Agent Mode remains active.
22.141 Turning Off Agent
Select the Agent control again when it shows:
Turn off Agent.
This stops:
22.142 Agent Mode and Transcription Share the Recognition Engine
The source starts the browser speech-recognition engine when either:
This creates an important relationship between those two functions.
22.143 Transcription Without AI Agent
You can turn on:
Transcription
without enabling Agent Mode.
In that case, recognized speech is still used to create and distribute transcript entries, but the Agent does not generate AI responses.
22.144 Agent Without the Transcript Panel
You can also enable Agent Mode without leaving the Transcript panel visible.
The AI speech engine can still listen for Agent interactions.
The Transcript panel controls whether the real-time transcription function is enabled/displayed independently of whether the Agent is active.
22.145 Important AI Agent Participation Limitation
The Agent's synthesized speech is generated through the local browser's speechSynthesis system.
The supplied source does not add the AI's synthesized voice as an outgoing WebRTC Audio track or broadcast the AI response through Meeting Data Channels.
Therefore:
The current source does not verify that other participants automatically hear or see your local AI Agent's response.
Treat Agent Mode as a participant-side AI assistant unless your live deployment testing proves additional behavior.
22.146 The AI Agent Is Not a Shared Meeting Bot in the Verified Source
Other participants are not automatically shown the AI Agent tile created on your browser.
Each participant's Room creates its own local Agent state.
This is another reason not to describe current Agent Mode as a centrally shared Meeting bot.
22.147 Fullscreen
On desktop, select:
Fullscreen.
AI Meeting requests browser Fullscreen for the application.
22.148 Exit Fullscreen
When Fullscreen is active, the tooltip changes to:
Exit fullscreen.
22.149 Participants, Chat, and Transcription Panels
The three right-side information panels are:
Only one of these is intentionally kept open at a time by the current Room controls.
22.150 Mobile Meeting Controls
On smaller screens, the primary toolbar keeps direct controls for:
while additional functions move into the:
More
menu.
22.151 Mobile More Menu
The Mobile More menu contains:
22.152 Leave
Select the red:
Leave
button to exit the current Meeting.
22.153 What Leave Does
The local Leave operation:
22.154 Leave Does Not Source-Verify “End Meeting for Everyone”
The current Meeting toolbar provides:
Leave
not:
End Meeting for Everyone.
The source does not establish a Host-only end-for-all control.
22.155 Leaving Does Not Automatically Mean the Meeting Record Is Ended
The local Leave handler does not itself call a Meeting API to mark the Meeting as ended.
It leaves the peer room and returns to the Lobby.
The backend rules governing Meeting-record status transitions are outside this User Operations bundle.
22.156 Media Tracks Are Stopped on Exit
When the Meeting Room unmounts, it also stops local and Screen Share media tracks.
This helps release your Camera and Microphone after leaving.
22.157 Current Camera/Microphone Device Selection
The media hook can enumerate:
and internally selects the first detected device when no selection exists.
22.158 No User-Facing Device Picker Is Rendered
Despite that underlying device support, the current Meeting Room does not expose a source-verified Settings/Device panel for manually selecting:
The Meeting Controls source does not render a Settings button or device-selection UI.
22.159 If the Wrong Camera or Microphone Is Used
Because a built-in device-picker UI is not currently established:
22.160 Speaker Selection Is Not Applied Through a Verified Meeting Control
Although the underlying media hook enumerates audio-output devices, the current Room source does not apply a selected Speaker to participant Video elements through a user-facing control.
Do not expect a current:
Speaker Output
menu inside AI Meeting.
22.161 Meeting Chat Persistence vs. Meeting Record Persistence
Keep these concepts separate:
Meeting Record
Source-verified as retrieved in Recent Meetings.
Meeting Chat
Current-session state and WebRTC Data Channel messages.
Live Transcript
Explicitly not stored permanently.
AI Agent Conversation
Local in-session AI Agent state.
Only the Meeting record has source-verified persistence in this chapter.
22.162 No Meeting Recording
The current AI Meeting source does not establish:
controls.
22.163 No Video Playback Archive
Recent Meetings are records, not recorded Meeting videos.
Selecting an active Meeting can reopen the live Room, but ended Meeting rows are not source-verified as replayable.
22.164 No Automatic Meeting Summary
AI Meeting does not currently produce a source-verified automatic:
at the end of the Meeting.
22.165 Meeting Notes Is the Separate Summary Workflow
The separate AI Meeting Notes Generator explicitly states:
“Transform meeting transcripts into structured, actionable notes with AI-powered analysis.”
That Tool accepts a supplied Transcript/Notes input and generates the structured analysis.
22.166 No Automatic AI Meeting → Meeting Notes Transfer Is Established
AI Meeting's Transcript is:
not stored permanently
and no current control exports it directly into Meeting Notes.
Therefore, this manual does not invent an automatic Meeting-to-Notes integration.
22.167 No Action-Item Extraction Inside the Live Room
Action Items belong to the separate Meeting Notes workflow.
The current live Room does not display an Action Items panel.
22.168 No Decisions Panel
Similarly, AI Meeting does not currently expose a live:
Key Decisions
panel.
22.169 No Follow-Up Generator Inside AI Meeting
The current Meeting Room does not contain a source-verified:
control.
Use the applicable PROS SUITE Tools separately when needed.
22.170 AI Meeting Token Catalogue Entry
The generic feature catalogue currently lists:
AI Meeting — 20 Tokens.
22.171 Important Token Qualification
Neither the AI Meeting Lobby nor core WebRTC Room contains an explicit client-side Token-deduction request for:
Therefore, the generic 20-Token catalogue value is not sufficient to prove a specific per-Meeting or per-minute billing rule.
22.172 AI Agent Can Invoke AI Usage
The optional AI Agent does call:
/api/ai/stream
using:
tool: "agent"
for spoken AI responses.
The server-side Token accounting for that AI request is outside the sanitized User Operations source.
22.173 Do Not Assume “20 Tokens per Meeting”
The current source does not verify that:
This manual will not assign billing behavior that the source does not establish.
22.174 AI Agent Model vs. Core Meeting Technology
The selected AI Model affects:
AI Agent responses
but does not determine:
Those are Meeting-system functions rather than Model-generation functions.
22.175 Transcription Does Not Require the Selected AI Model in Current Source
Live Transcription uses browser Speech Recognition.
When Agent Mode is inactive, recognized text can be sent to the Transcript without generating an AI response.
Therefore, Live Transcript and AI Agent are related but not identical functions.
22.176 Recommended Meeting Creation Workflow
AI MEETING
↓
NEW MEETING
↓
Optional:
MEETING TITLE
↓
CREATE & JOIN
↓
Allow:
CAMERA + MICROPHONE
↓
Review:
PRE-JOIN PREVIEW
↓
Set:
MUTE / CAMERA
↓
JOIN MEETING
↓
INVITE
↓
Share Meeting link with authorized participants.
22.177 Recommended Join Workflow
AI MEETING
↓
JOIN MEETING
↓
Paste:
MEETING ID OR LINK
↓
Press Enter
or:
JOIN MEETING
↓
Allow Camera/Microphone
↓
Review Preview
↓
Choose initial Mute/Camera state
↓
JOIN MEETING
22.178 Recommended In-Meeting Workflow
Once connected:
22.179 Recommended Transcription Workflow
JOIN MEETING
↓
Select:
TRANSCRIPTION
↓
Live panel displays:
Waiting for speech...
↓
Speak normally
↓
Recognized phrases appear with:
↓
Review during the Meeting
↓
Remember:
Transcript is not stored permanently
22.180 Recommended AI Agent Workflow
JOIN MEETING
↓
Choose desired Header AI Model if necessary
↓
Confirm applicable provider API Key
↓
Select:
AGENT MODE
↓
Agent becomes:
LISTENING
↓
Speak a question
↓
Agent becomes:
PROCESSING
↓
AI response generated
↓
Agent becomes:
SPEAKING
↓
Browser speaks the response
↓
Agent resumes:
LISTENING
22.181 Example AI Agent Questions
Examples of concise Meeting-side questions include:
Summarize the main point I just explained in one sentence.
What are three risks we should consider?
Give me a concise way to explain this proposal.
What question should I ask next?
Because Agent Mode maintains its local conversation context, follow-ups can build on earlier Agent exchanges during the active Agent session.
22.182 Do Not Assume Agent Heard All Participants
The Agent listens through the local browser's speech-recognition environment.
The source does not establish centralized capture/transcription of a mixed Meeting audio stream for AI analysis.
Do not assume it has a complete authoritative transcript of everything every participant said.
22.183 AI Agent Is Not Automatic Meeting Minutes
Agent Mode provides conversational assistance.
It is not the same as:
AI Meeting Notes Generator.
22.184 Best Practice — Name Meetings
Although the Title is optional, use a meaningful Name when Meetings matter operationally.
This improves Recent Meetings identification.
22.185 Best Practice — Test Media Before Joining
Use the Pre-Join screen to confirm:
before entering.
22.186 Best Practice — Check Browser Permissions in Advance
For important Meetings, confirm your browser already has permission for:
before the scheduled Meeting starts.
22.187 Best Practice — Use a Secure Browser Context
Media devices require HTTPS or an appropriate secure local environment.
Use the normal secure AIUNIFY PROS SUITE website rather than an insecure copied address.
22.188 Best Practice — Share the Meeting Link Only with Intended Participants
Meeting links contain the Meeting ID.
Distribute them through an appropriate channel to people who should participate.
22.189 Best Practice — Remember Authentication Is Required
Tell invitees that the current PROS SUITE source requires an authenticated AI Meeting session.
Do not promise anonymous one-click guest access.
22.190 Best Practice — Mute When Not Speaking
Especially in larger Meetings, muting reduces unwanted:
Use the toolbar's Mute control.
22.191 Best Practice — Use Screen Share for Visual Collaboration
When discussing:
Screen Share gives participants a shared visual reference.
22.192 Best Practice — Remember Screen Audio Is Not Included
If you are presenting a Video or audio demonstration, the current screen-share source does not request system audio.
Plan accordingly.
22.193 Best Practice — Use Chat for Short Links or Notes
Meeting Chat is appropriate for quick Text such as:
Do not rely on it as a permanent Meeting record.
22.194 Best Practice — Do Not Rely on Transcript Persistence
The interface explicitly says transcripts are not stored permanently.
If an important Meeting requires an official transcript, AI Meeting's current Live Transcript should not be treated as a durable archive.
22.195 Best Practice — Obtain Appropriate Consent
Before using transcription or other Meeting processing, follow applicable organizational policies and legal requirements regarding participant notice or consent.
AI Meeting's technical ability to transcribe does not replace those responsibilities.
22.196 Best Practice — Use Agent Mode Intentionally
Agent Mode continuously listens while active.
Turn it on when you actively want an AI assistant and turn it off when you do not.
22.197 Best Practice — Keep AI Agent Questions Concise
The AI Agent is specifically instructed to generate concise spoken answers.
Questions that can be answered conversationally fit the current experience best.
22.198 Best Practice — Verify AI Agent Answers
The AI Agent remains generative AI.
Do not treat spoken responses as automatically verified facts, particularly for:
22.199 Best Practice — Use the Separate Meeting Notes Tool for Structured Analysis
If you need:
use the dedicated Meeting Notes workflow with an appropriate transcript or notes source.
22.200 Troubleshooting — Cannot Access AI Meeting
Check:
Both the Lobby and Room are protected.
22.201 Troubleshooting — Invite Link Goes to Login
This is expected for an unauthenticated participant in the current source.
Meeting-room routes require authentication.
22.202 Troubleshooting — Invitee Has an Account but Still Cannot Enter
The Meeting Room also uses:
FeatureGuard featureId="ai-meeting".
The invitee must therefore have access to the feature as well.
22.203 Troubleshooting — Invalid Meeting ID
If you receive:
“Invalid meeting ID or link. Please check and try again.”
review the copied link or ID.
22.204 Troubleshooting — Unable to Join
Read the Error displayed beneath:
Unable to Join.
Then select:
as appropriate.
22.205 Troubleshooting — This Meeting Has Ended
The Meeting service reported that the Meeting status is ended.
The current Room refuses entry.
Create or request a new Meeting.
22.206 Troubleshooting — Camera/Microphone Permission Denied
Open your browser's Site Permissions and allow:
Then reload the Meeting.
The exact current permission Error directs users to their browser settings.
22.207 Troubleshooting — No Camera or Microphone Found
Confirm:
Then retry.
22.208 Troubleshooting — Wrong Camera Is Selected
No user-facing device picker is currently verified.
Adjust the browser or operating system's preferred Camera and reload.
22.209 Troubleshooting — Wrong Microphone Is Selected
Use the same approach:
then reload the Meeting.
22.210 Troubleshooting — Participant Has No Video
Possible source-aligned causes include:
The participant tile falls back to an avatar-style display when usable Video is unavailable.
22.211 Troubleshooting — Peer Disconnects
The WebRTC layer monitors connection state.
If a peer connection fails, the source can attempt an ICE restart and can eventually remove a failed/closed peer from the local participant map.
22.212 Troubleshooting — Screen Share Fails
Possible reasons include:
Try again using:
Share screen.
22.213 Troubleshooting — My Screen Share Stopped When Someone Else Shared
This is expected current behavior.
When another participant begins a Screen Share while yours is active, the local Room stops your share.
22.214 Troubleshooting — Participants Cannot Hear Video Audio from My Screen
Current Screen Share requests:
audio: false.
System audio sharing is not established.
22.215 Troubleshooting — Meeting Chat Disappeared
Opening:
closes Meeting Chat.
Select Chat again to reopen it.
22.216 Troubleshooting — Chat Is Gone After Meeting
Current Meeting Chat is not source-verified as persistent.
Do not depend on it for post-Meeting records.
22.217 Troubleshooting — Transcription Shows “Waiting for speech...”
Possible reasons include:
22.218 Troubleshooting — Transcription Does Not Work in My Browser
The source depends on:
Try a compatible modern browser if Live Transcript is required.
22.219 Troubleshooting — Transcription Is Using English
The current speech-recognition language is fixed to:
en-US.
There is no current language selector in AI Meeting.
22.220 Troubleshooting — I Cannot Download the Transcript
No Transcript Download control is source-verified.
The interface explicitly says Transcript data is not stored permanently.
22.221 Troubleshooting — AI Agent Shows Connection Error
The Agent can enter:
Connection Error
when its speech-recognition or AI processing path encounters an Error.
Check:
22.222 Troubleshooting — AI Agent Does Not Hear Me
Check:
22.223 Troubleshooting — AI Agent Stops Listening While Answering
This is expected.
The Agent temporarily stops recognition while:
then resumes listening afterward.
22.224 Troubleshooting — Other Participants Cannot Hear My AI Agent
The current source does not transmit synthesized Agent speech as a WebRTC Audio track.
Therefore, other participants hearing your Agent is not source-verified.
This is an implementation limitation, not necessarily a Microphone problem.
22.225 Troubleshooting — Other Participants Cannot See My AI Agent Tile
The Agent tile is generated by each participant's local Meeting UI.
The source does not broadcast Agent activation as a central Meeting participant.
22.226 Troubleshooting — No Meeting Summary Appeared After Leaving
This is expected.
AI Meeting does not source-verify an automatic post-Meeting Summary.
Use the separate Meeting Notes Tool when structured analysis is needed.
22.227 Troubleshooting — Cannot Find Record Meeting
No Meeting Recording control is established in the current source.
22.228 Troubleshooting — Cannot Remove Another Participant
No Host removal control is established.
The current Participants panel is primarily informational.
22.229 Troubleshooting — Cannot End Meeting for Everyone
The current toolbar contains:
Leave
rather than a Host:
End Meeting for Everyone
function.
22.230 Troubleshooting — Cannot Schedule a Meeting
The current AI Meeting Lobby creates instant Meetings.
It does not establish future scheduling controls.
22.231 AI Meeting Creation Checklist
Before creating:
22.232 Invite Checklist
Before inviting another person:
22.233 Pre-Join Checklist
Before selecting Join Meeting:
22.234 In-Meeting Control Checklist
You should now know how to use:
22.235 Transcription Checklist
Before relying on Live Transcript:
22.236 AI Agent Checklist
Before using Agent Mode:
22.237 End-of-Meeting Checklist
Before leaving:
22.238 What AI Meeting Currently Supports
The source verifies subscriber-facing functionality for:
Lobby
Pre-Join
Live Meeting
AI Functions
22.239 What AI Meeting Does Not Currently Establish
The supplied user-facing source does not establish:
Do not document these as active user features unless a future source/UI release verifies them.
22.240 AI Meeting vs. Meeting Notes at a Glance
AI Meeting
Conduct the Meeting
Meeting Notes
Analyze supplied Meeting text
22.241 AI Meeting at a Glance
The complete current workflow is:
AI MEETING
↓
Choose:
NEW MEETING
or:
JOIN MEETING
↓
For New:
Meeting Title (optional)
↓
CREATE & JOIN
or for existing:
MEETING ID OR LINK
↓
PRE-JOIN
↓
Allow:
CAMERA + MICROPHONE
↓
Choose:
MUTE / UNMUTE
CAMERA ON / OFF
↓
JOIN MEETING
↓
Live controls:
MICROPHONE
CAMERA
SCREEN SHARE
PARTICIPANTS
CHAT
TRANSCRIPTION
AGENT MODE
FULLSCREEN
INVITE
↓
Optional AI workflow:
LISTENING
→ PROCESSING
→ SPEAKING
↓
When finished:
LEAVE
↓
Return to:
AI MEETING LOBBY
22.242 Chapter Summary
AIUNIFY PROS SUITE AI Meeting is a live browser-based Meeting system built around WebRTC peer-to-peer communications.
The Lobby describes the system as:
“Create or join secure video meetings instantly. Crystal-clear audio and video powered by WebRTC peer-to-peer technology.”
Subscribers can create an instant Meeting using:
New Meeting → Meeting Title (optional) → Create & Join
or enter an existing session through:
Join Meeting → Meeting ID or Link.
The current Meeting Room requires authenticated PROS SUITE access, including when entering from a shared Meeting link. Anonymous guest access is therefore not source-verified.
Before joining, AI Meeting requests Camera and Microphone permission and displays a local Preview where the user can adjust their initial Mute and Video state. Current permission errors explicitly cover denied access, missing devices, and unsupported/insecure browser contexts.
Inside the Meeting, subscribers can use: