Set up how your agent looks at each call after it ends. Post-call analysis gives you organized data from every conversation. 📊
Post-call statuses label the result of each call. The agent picks the best status based on how the conversation went.
| Status | Description |
|---|---|
| Completed | The call went well and got the result you wanted ✅ |
| No Answer | The client didn't pick up 📵 |
| Voicemail Left | The agent left a voicemail message 📭 |
| Invalid or Disconnected Number | The phone number doesn't work ❌ |
| Wrong Party Reached | The agent reached someone else, not the right client |
| Information Collected | Got the info that was needed 📝 |
| Documents Requested | Asked the client to send documents 📄 |
| Escalated to Staff | The call was sent to a real team member 🙋 |
| Follow-Up Required | The client needs another call later 🔄 |
| Not Qualified | The client doesn't meet the requirements |
You can add, change, or remove statuses to fit your needs:
TIP
Post-call statuses can kick off actions in campaign workflows! For example, you can move a contact to a different list, update the CRM, or send a follow-up message based on how the call went. 🔗
Post-call variables are pieces of info that the AI pulls out of each conversation. You can use this data in automations and reports.
| Variable | Type | Description |
|---|---|---|
| Call Summary | String | A short recap of what happened on the call 📝 |
| Call Successful | Boolean | Did the call reach its goal? (yes or no) |
| Voice Call Opt-Out | Boolean | The client asked to stop getting calls 🚫 |
| Wrong Number Detected | Boolean | The call went to the wrong person |
| Disqualified | Boolean | The client opted out or doesn't qualify |
| Form Disqualified by Task | Boolean | A specific form answer caused disqualification |
| Converted | Boolean | The client did what you wanted (e.g., agreed to transfer, finished intake) 🎉 |
Add your own variables to pull out specific info from calls:
| Setting | Description |
|---|---|
| Name | A unique name for this variable |
| Description | Tell the AI what to look for |
| Type | The kind of data: String, Boolean, or Enum |
| Type | Description | Extra Settings |
|---|---|---|
| String | Any text pulled from the conversation | You can add examples to help the AI |
| Boolean | A yes-or-no answer | — |
| Enum | One choice from a list you set up | You must add the list of choices |
String variable:
injury_typeBoolean variable:
has_attorneyEnum variable:
urgency_levelPick the AI model used to look at calls after they end:
| Model | Description |
|---|---|
| GPT-4.1 | High quality, a bit slower |
| GPT-4.1 Mini | Good mix of quality and speed ⚖️ |
| GPT-4.1 Nano | Fastest and cheapest 🚀 |
| Claude 4.5 Sonnet | Another high-quality option |
| Claude 4.5 Haiku | A fast alternative |
| Gemini 2.5 Flash | Google's speedy model ⚡ |
INFO
This model only affects what happens after the call — not the live conversation. Pick a faster model if you want your call data ready quickly. ⏩
Statuses and variables from calls flow into other parts of Kayse AI:
| Integration | How It's Used |
|---|---|
| Campaign Workflows | Trigger actions based on call results (e.g., send a follow-up, update CRM) |
| Campaign Reports | Add post-call data to scheduled CSV reports 📄 |
| Analytics | Track results across all your campaigns 📊 |
| Automations | Use call data to start automatic actions ⚙️ |
| CRM Sync | Match post-call variables to fields in your CRM 🔄 |