Why Monitor exists
Two jobs sit on every team lead’s plate: making sure the work going out the door is good, and making sure the people doing it are getting better. Monitor is built for both — quality control on one shoulder, training and retraining on the other.
1The two jobs Monitor does
— quality on one side, growth on the other
Quality control
spot-check & review
Catch what matters before it ships
Sample live conversations, watch tone in real time, and flag the ones that need a second pair of eyes — escalation moments, tricky de-escalations, anything where the AI handed off to a human and the handover deserves a look. Quality control here isn’t an audit at the end of the week; it’s a continuous, lightweight habit you build into the day.
Training & retraining
onboarding, coaching, follow-up
Turn real conversations into coaching material
New teammates need a watchful eye in their first weeks. Experienced teammates sometimes need a refresher on a specific topic, channel, or product line. Monitor gives you a way to track exactly the people and queues you’re focusing on right now — without losing sight of everything else.
Not every leader needs to see every conversation. A lead on the billing team doesn’t need voice queue chatter. A new agent’s mentor wants to watch just that agent’s work for the first week. Monitor lets you carve out the exact slice of the operation you’re responsible for, and ignore the rest.
Building a view from filters
A view is a saved combination of filters. Stack as many as you need to narrow the surface to exactly the conversations you want to watch — then save it, name it, and it’s waiting for you next time.
1The filters you can stack
— combine freely; views save the whole stack
◉
Team
One or many customer service teams
⌬
Agent
Specific individuals you’re tracking
⤳
Channel
Email, chat, voice, SMS, WhatsApp, Messenger
#
Tag
Any tag your team applies to conversations
◇
AI status
AI-managed, human-managed, or handed-off
⏱
Time range
Limit to a window — today, last hour, this shift
🌐︎
Language
Detected or translated language
⊕
Specific IDs
Pin individual conversations into a view
Where stacked filters earn their keep ↓
2Three views worth saving
— recipes that translate directly into training and quality work
Recipe 1
Newly onboarded teammates
Filter by Agent — the names of the people in their first weeks — and leave everything else open. You see only their conversations, on every channel they’re working. Watch in the morning, flag anything worth coaching on, then walk through the flagged set in your one-on-one. The same view stays useful for as long as the cohort needs close attention; remove names as they graduate.
Recipe 2
Individuals who need a follow-up
After a coaching session you want to verify the lesson stuck. Build a view with the agent’s name plus the relevant Tag or Channel — “refunds,” “voice,” whatever the topic was — and check it for a week. Targeted, time-bound, then retire the view.
Recipe 3
An entire team under closer watch
When a team is rolling out a new product, learning a new tool, or recovering from an incident, narrow Monitor to Team + the relevant Tag and watch the queue together. Same view works for surge management — pin it during a busy period, set it aside when things calm down.
Shared filter presets travel with the team. Save a view at the tenant level and every supervisor in your organisation can use the same lens. This is how leadership keeps everyone looking at the same picture — fewer arguments about which numbers are real, more time spent acting on them.
Pin a view as your default. Mark any saved view as your default and that’s where Monitor opens — the curated slice you’re responsible for, ready when you are. Switch between saved views as the day demands; the default is the one waiting for you in the morning.
Working in real time
Monitor is a live view, not a report. Cards update without a refresh, the conversation panel opens to the right, and you choose how dense the surface should feel.
1Card density
— pick how much you want to see at once
▤
Compact
Minimum information, maximum cards visible
▥
Normal
Balanced — the everyday default
▦
Detailed
Full context — message previews and metadata
And underneath, the live updates that make it worth watching ↓
2What updates on its own
— no refresh button required
Cards refresh on a short cadence
Active conversations update every few seconds — when a new message arrives, when status changes, when an agent picks up a card, you see it without lifting a finger. Resolved conversations stay visible for a configurable window so you can review them after the fact.
Typing indicators surface live work
When an agent starts composing a reply, Monitor shows it. You can tell the difference between a queue that’s being worked and one that’s been parked — at a glance, without opening the card.
3The conversation panel
— review without leaving Monitor
Click a card, the full conversation opens to the right
The same surface you’d see in the inbox — message timeline, contact details, event history, side conversations, internal comments, Sidekick suggestions — opens in a panel beside your view. Read it, flag it, drop a comment, start a Whisper, and close it. You never lose the queue you’re watching.
The retention window is yours to set. Done conversations stay visible for the period you configure — long enough for a quality pass, short enough that the surface doesn’t become a graveyard. Adjust per view; one number per surface.
Two ways to step in
Watching is half the job. The other half is acting on what you see — sometimes after the fact, sometimes in the moment. Monitor has a tool for each: Flagging for review later, Whisper for help right now.
1Flagging — review later
— async coaching, with the context preserved
Mark a conversation, leave a comment, walk through it later
When you spot something worth talking about — a graceful save, an awkward escalation, a tone that drifted — flag the conversation. A red flag appears on the card and follows the conversation through the inbox, and you attach a short comment explaining what you noticed: “Great de-escalation here,” “Check the tone in message three,” “This should have been escalated sooner.” Edit the comment as the situation evolves, add follow-up notes, or remove the flag once the review is closed out.
The end-of-week ritual. Flag conversations during the day as you spot them. At week’s end, walk the flagged set with the agent in your one-on-one — the comments you left are the agenda. Specific, actionable, anchored in real work.
When the moment can’t wait, the second tool ↓
2Whisper — help right now
— private real-time coaching during a live conversation
Private
customer never sees it
A side channel just for the two of you
Whisper opens a private thread between the supervisor and the agent, attached to the live conversation but invisible to the customer. Anything you send lands in the agent’s view as a coaching message — not in the customer’s inbox, not in the conversation transcript.
Real time
typing indicators on both sides
Conversation, not commentary
Whisper is a two-way thread. Send a nudge, the agent replies, you go back and forth. Typing indicators show when either party is composing — so the agent knows help is on the way and you know they’re reading.
Initiated by leads
admins & owners
The supervisor opens the thread
Whisper is initiated from Monitor by a team lead, admin, or owner — not by the agent on their own. Once it’s open, both sides can write freely; the agent doesn’t need to wait to be addressed.
Recorded
attached to the conversation
Every Whisper stays with the case
The thread is preserved alongside the conversation. You can revisit what was said and when, use it in a later coaching session, or cite it when reviewing how a difficult case was handled.
The line between Flag and Whisper is timing. Flag is for the review you’ll have on Friday; Whisper is for the moment that can’t wait until Friday. Use Whisper when an active conversation is about to go sideways and the agent needs a hand — a script, a discount limit, a reminder of policy. Use Flag for everything else.
