Explorer

The Explorer is where you go to actually understand your conversations. It lets you slice them by any dimension, narrow to the segment you care about, break that down further, watch it trend over time, and — crucially — read the real conversations behind every number. It's the difference between "sentiment dropped" and "sentiment dropped because of shipping delays in İzmir."

Three ways to slice

Explorer lets you look at conversations through three kinds of lens, and you can combine them freely:

• What was said — the classification dimensions: sentiment, request type, urgency, topic, and any custom ones you've added.
• About the conversation — details like the detected language.
• About the visitor — their device, browser, country, and city.

Pick any of these and Explorer shows you the breakdown — how your conversations distribute across its values.

The rhythm of exploring

Exploring is a loop of narrow, then break down. It goes like this:

1. Start broad. Pick a dimension — say Sentiment — and see the split across positive, neutral, and negative.
2. Narrow to a segment. Click Negative to filter down to just the unhappy conversations.
3. Break it down again. Now split that segment by another dimension — Topic — to see what the negative conversations are actually about.
4. Watch it move. Chart the segment's trend over time to see whether it's getting better or worse.
5. Read the proof. Open the actual conversations in the segment and read what customers said.

That last step is the one that turns numbers into understanding. A chart tells you shipping complaints are up; the conversations tell you why.

Stacking filters

You're never limited to one filter. Explorer lets you stack them to home in on exactly the slice you care about:

Negative sentiment and high urgency and topic is "delivery" and country is Türkiye

The more you stack, the sharper the picture — until you're looking at precisely the conversations that matter for the question you're asking.

Save what you'll come back to

When you've built a segment you'll want to revisit — "angry delivery complaints from returning customers" — you can save it. From then on it's one click away, so you can track that exact slice over time instead of rebuilding the filters each visit.

From exploring to fixing

Explorer is built for the journey from symptom to cause to action. You spot a worrying number, narrow until you understand it, read the conversations to know what's really going on — and then you can act: improve a flow, add knowledge, or fix a process.

When the issue is that the agent simply couldn't answer something, head to Knowledge Gaps — it surfaces exactly those moments and helps you close them. And the patterns Explorer makes you hunt for by hand, Findings often spots for you automatically.