Auto summary
A plain language summary of what the visitor wants and what's already been resolved, ready before an agent opens the conversation.
Co-pilot
An analyst riding along on every conversation
Co-pilot reads every conversation as it happens tagging intent, flagging what's missing, and logging every AI decision, so your team opens a case already knowing the story.
Connect your web chat & social account in under 3 minutes.
Available today
Understands the conversation before your team does
No setup, no prompting. Co-pilot classifies intent, drafts a summary, and tags the lead the moment a conversation starts.
Intent & missing fields
Detects what the visitor is trying to do book, ask, complain and flags exactly what's still missing, like a phone number or an order ID.
Lead & pipeline tagging
Qualifies each conversation as hot, warm, or cold, and moves it through Open, Won, or Lost as the conversation develops.
How it works
A running log of what the AI decided, and why
Every escalation, tag, and status change is written to an activity log with an actor and a timestamp, so nothing an agent picks up is a black box.
Full activity trail
Every AI action, escalation, metadata update, tag change is logged with a timestamp and before and after values, visible to any agent who opens the conversation.
Verified visitor data
Attributes passed from your own site role, account ID, organization sit alongside the conversation, verified and ready to use.
Escalation with context
When Co-pilot can't resolve something, it hands off with the intent already classified and the missing information already named.
Where it pays off
From reading the thread to already knowing the story
P / 1
Agents open conversations cold.
Co-pilot's summary means an agent claiming an escalated chat starts with what happened and what's needed, not a scroll through the transcript.
P / 2
Follow ups stall on missing information.
Missing field detection flags a phone number or order ID the moment it's needed, instead of three replies later.
P / 3
Lead status lives in someone's head.
Qualification and pipeline stage are set from the conversation itself, so Open, Won, and Lost are decisions the record already reflects.
Stack comparison
Reading conversations today vs. with Fellix Co-pilot
Picking up a conversationStatus quoAn agent reads the whole thread to understand what's going on.FellixA summary and detected intent are ready before they open it.
Missing informationStatus quoDiscovered mid-reply, or missed entirely.FellixFlagged the moment it's needed, not three replies later.
EscalationsStatus quoHanded off with a note, if there's time to write one.FellixHanded off with intent and missing fields already attached.
Lead qualificationStatus quoSet manually, if at all, after the fact.FellixTagged hot, warm, or cold as the conversation happens.
Audit trailStatus quoWhatever's in the CRM notes, written after the fact.FellixEvery AI action logged with actor, timestamp, and changes.
Trusting the handoffStatus quoDepends on who wrote the note, and whether they wrote one.FellixTraceable — any agent can see what the AI concluded and why.
Integrations
Connect Fellix to the tools your data already lives in
Fellix reads and writes to your CRM, storefront, calendar and helpdesk — so every conversation starts with real customer context, and every outcome lands back in the stack your team already works in.
Built to work together