Sessions
The Sessions page gives you a clear view of all past conversations between users and your AI Agent. It’s where insights turn into improvements — helping you understand what’s working, where users drop off, and how to make your AI more persuasive and human-like over time.
Each session captures a complete transcript, letting you see exactly how your AI responded and what led to a successful (or failed) outcome.
When to Use
Use the Sessions dashboard when you want to:
- Review real user conversations after going live.
- Spot where users lose interest or get confused.
- Identify unanswered questions (Missing Knowledge).
- Examine sessions that successfully achieved a goal.
- Optimize tone, flow, and user experience for higher conversions.
How It Works
Accessing Sessions
On your left-hand navigation, click Sessions.
You’ll land on a list of all conversations, sorted by most recent first.
What You’ll See
Each session card provides quick-glance details:
- User name & status (Active / Inactive)
- Session duration and message count
- Sentiment indicators (for flagged negative interactions)
- Goal completion icons (for achieved objectives)
Click any session to open the full chat transcript. You can then replay or scroll through the interaction to analyze how your AI responded, which Journey blocks were triggered, and what data was collected.
Filters & Views
The Sessions dashboard includes powerful filters so you can focus your analysis:
Filter | Description |
|---|---|
All Channels | Choose between Web Chat, SMS, Email, or other connected channels. |
All Sentiment | Filter by conversation tone — Positive, Neutral, or Negative. |
Has Responses | Show only conversations where users replied. |
Name / Email / Phone | Find sessions tied to specific users or leads. |
All Types | Filter by conversation type (e.g., lead capture, support). |
Agents | Focus on sessions handled by a specific AI Agent. |
Goals | View sessions where goals were triggered or completed. |
Journey Blocks | Identify which conversational block the user reached or exited. |
Forms | Filter sessions that included form submissions. |
Status | Sort by open, active, or closed sessions. |
Tags | Group sessions by campaign or workflow tag. |
💡 Tip: Use multiple filters together (for example: “Web + Negative Sentiment + Goal Not Met”) to pinpoint specific areas for optimization.
How to Use It for Optimization
Once you have enough live sessions, use them to refine your AI performance strategically:
Fix Negative Experiences
- Filter by Negative Sentiment.
- Identify where the tone or logic went wrong.
- Update your Guardrails or Persona to improve empathy and clarity.
Fill Knowledge Gaps
- Filter by Missing Knowledge sessions.
- Add the missing details to your Knowledge Collection or Playbooks.
Replicate Success Patterns
- Review sessions that achieved Goal Conversions.
- Look for what phrasing or Journey flow drove success.
- Reuse these elements across similar Journeys or agents.
Audit AI Reasoning
- Click the 🤖 Robot Head Icon inside a session
- Inspect how your AI decided on responses, which knowledge it used, which block it used, and what conditions triggered actions.
Examples / Templates
Example Use Case 1: Improving Onboarding Chat
- Filter by Missing Knowledge for your onboarding AI.
- Notice multiple users asking “How do I reset my password?”
- Add an FAQ entry or snippet to your Knowledge Collection.
Example Use Case 2: Conversion Audit
- Filter sessions with Goal = Demo Booked.
- Compare how your AI handled pricing or objections.
- Replicate the successful tone and flow in future campaigns.
Best Practices
- Review sessions weekly during your first month after launch.
- Tag recurring issues and address them through Knowledge updates.
- Correlate with Leads to see how high-quality sessions translate into conversions.
- Use the Robot Head tool to understand AI behavior — don’t just fix responses, fix the logic.
- Always link insights back to improving Journeys, not just answers.