Chloe is the AI assistant built into Gruvi. She is your customer support agent, and she directs every other agent working in your workspace.
She is one half of the AI in Gruvi. The other is the empathy engine, which works out how each customer is feeling and changes what Chloe does about it.
Who Chloe is
Treat Chloe as a new colleague rather than a feature. She reads your conversations, learns from your material, and gets better the more you correct her.
Like a new colleague, she starts out knowing nothing about your business. What she knows is what you have given her.
What Chloe can do
- Draft a reply for you to check and send.
- Rewrite a draft you have already written.
- Summarise a long thread so you can pick it up quickly.
- Reply to customers on her own, if you let her.
- Watch conversations quietly and build up knowledge without saying anything.
- Answer your questions about your own workspace, in the Talk with Chloe area.
How many of these are switched on depends entirely on her mode. That is covered in Autopilot, On demand, and Learning modes explained.
Where Chloe gets her answers
Chloe does not invent answers about your business. She draws on three things you control:
- Knowledge sources
- Your website, uploaded documents and articles you have written. She distils facts from these and keeps them in mind for every reply, each one backed by a source article.
- Do's and don'ts
- Plain-language rules she follows in every reply. The list is deliberately capped so that every rule stays sharp.
- Guided answers
- Specific answers to specific questions, created when you fix one of her replies, or added by hand.
How Chloe sounds
You set her tone on a slider running from Technical to Empathetic, with five stops: Technical, Leans technical, Balanced, Leans warm, and Empathetic.
Both ends are equally capable. They sound different. A preview underneath shows the same reply written in the tone you have chosen, so you can hear the difference before committing.
When Chloe steps back
Chloe hands a conversation to a human when an escalation rule matches. When that happens she stops replying, leaves a draft for your team, and marks the reply Flagged for a human in the feed.
Four rules are on by default:
- Billing disputes, refund requests, or payment problems.
- Legal threats or mentions of legal action.
- The customer explicitly asks for a human, or severe frustration continues after the assistant has failed to help.
- Account closure or data deletion requests.
You can switch any of these off, and add your own in plain language. Chloe applies them in whatever language the customer writes in.
What a handover looks like
It is worth knowing what your customer sees, because it is not an error message. Chloe says plainly that she is bringing a person in, and names what she is bringing them in for. She says something like "I'm bringing a human teammate into this conversation to help with your question about the free plan's document limit."
On your side, three things happen. The thread moves into the Inbox queue. A headphones icon appears on its card, meaning Chloe handed this over to the team. And a line appears in the conversation itself recording the handover.
Where to find Chloe
Chloe has her own section in the left-hand sidebar, under Agents. Clicking her name gives you three areas:
- Talk with Chloe
- Ask her about this workspace, or tell her what to set up.
- Training
- Rehearse with her, set rules, and manage guided answers.
- Settings
- Her mode, tone, knowledge and handoff rules.
Turning Chloe off
At the bottom of her settings there is a Turn off agent button. It shuts Chloe down entirely: no replies, no drafts, no learning and no background analysis.
It is not permanent. You can turn her back on whenever you like, and nothing is lost.
For a single conversation you do not need anything so drastic. Click Pause on the Chloe auto-replies here bar and she stops on that thread alone. ---