## Best Visualization Approach Retell AI has built-in Performance Analytics (described as "Mixpanel for AI Voice Agents") [(5) ]() that provides: - Call success rate, transfer rate, call counts - Latency variations monitoring - A/B test comparison across agents - Custom chart creation - Deep dive into programmatic calls This analytics dashboard became available for free and includes advanced filtering capabilities to spot high-latency calls and monitor performance . For custom dashboards, the List Calls API returns comprehensive data including [(1) ]() : -

transcript

,

recording_url

,

duration_ms

-

latency

metrics (e2e, asr, llm, tts, knowledge_base) -

call_analysis

with summary, sentiment, success status -

call_cost

breakdown by product You can export call history data using selected filters and column settings [(5) ]() , which mirrors the call history table for external visualization tools.

### Answer Based on the available sources, here are the key endpoints and methods for building a monitoring dashboard: ## Core Monitoring Endpoints List Calls API - Retrieve call details with filtering and pagination [(1) ]() :

javascript
import Retell from 'retell-sdk';
const client = new Retell({
 apiKey: process.env['RETELL_API_KEY'],
}) ;
const callResponses = await client.call.list() ;

This endpoint supports POST requests to

/v2/list-calls

with extensive filter criteria including : -

call_id

,

agent_id

,

call_status

,

call_type

-

direction

(inbound/outbound) -

user_sentiment

,

call_successful

-

start_timestamp

with upper/lower thresholds -

metadata

and

dynamic_variables

filtering Get Call API - Retrieve individual call analysis programmatically [(2) ]() Conversation Flow API - Get conversation flow details [(3) ]() :

javascript
const conversationFlowResponse = await client.conversationFlow.retrieve('conversation_flow_id') ;

## Real-Time Data via Webhooks Set up webhooks to receive real-time call events [(4) ]() :

javascript
app.post("/webhook", (req, res) => {
 const {event, call} = req.body;
 switch (event) {
 case "call_started":
 console.log("Call started event", call.call_id) ;
 break;
 case "call_ended":
 console.log("Call ended event", call.call_id) ;
 break;
 case "call_analyzed":
 console.log("Call analyzed event", call.call_id) ;
 break;
 }
 res.status(204) .send() ;
}) ;

​ To learn more, see the sources I considered: [[1] List Calls](https://docs.retellai.com/api-references/list-calls) [[2] Consume the analysis data](https://docs.retellai.com/features/post-call-analysis-consumption) [[3] Get Conversation Flow](https://docs.retellai.com/api-references/get-conversation-flow) [[4] Setup guide](https://docs.retellai.com/features/register-webhook) [[5] Platform Changelogs | Retell AI](https://www.retellai.com/changelog) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

### Answer Based on the available sources, live transcription capability depends on the call type: For Web Calls: Live transcription is available [(1) ]() . The platform changelog confirms this feature was added for WebCall implementations. For Phone Calls: According to a Discord discussion, monitoring live phone calls with real-time audio and transcription is not currently possible for standard phone calls [(2) ]() . However, there is a potential workaround: if you use a Twilio phone number bound to your Retell agent and employ the "dial to SIP" method, this approach is viable for streaming back audio and transcript . The limitation exists because the deprecated audio websocket () was the previous method for accessing real-time audio streams . For building your dashboard with live monitoring capabilities, you'll need to: - Use web calls if you need guaranteed live transcription support - Consider the Twilio + SIP integration method if you must use phone calls - Otherwise, rely on post-call transcripts available through the List Calls API and webhooks for near-real-time updates

​ To learn more, see the sources I considered: [[1] Platform Changelogs | Retell AI](https://www.retellai.com/changelog) [[2] Monitor live phone calls (not web calls) and receive the real time audio + transcription](https://discord.com/channels/1186424376300679209/1469419908692705487) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

### Answer Based on the available sources, here's a realistic feasibility breakdown: ## RETELL DATA YOU CAN DIRECTLY USE ✅ Fully Supported: - Call history, transcripts, recordings, summaries, sentiment analysis, call duration/status, agent filtering, latency metrics, and call costs are all available through the List Calls API [(1) ]() - Webhook events support

call_started

,

call_ended

,

call_analyzed

, and

transcript_updated

[(2) ]() ## REALTIME OPERATIONAL FEATURES ⚠️ Partially Supported: - **Near realtime transcript updates**: Available via

transcript_updated

webhook event, which triggers on turn-taking transcript updates plus a final update when the call ends - **Operational analytics dashboard**: Retell provides a built-in Analytics dashboard with customizable charts, global filters, and real-time data updates [(3) ]() - **Latency telemetry**: Available in the

latency

object from List Calls API, including e2e, asr, llm, tts, and knowledge_base metrics [(1) ]() Not Documented: - Live active call monitoring with real-time audio/transcript for phone calls - AI/User speaking indicators during calls - Event stream logs (beyond webhook events) - Silence/interruption alerts ## CUSTOM FEATURES YOU BUILD ✅ Highly Feasible: - **Conversation intelligence layer**: Build on top of transcript data and call analysis - **AI quality scoring**: Use

call_analysis

data including

call_successful

,

user_sentiment

, and custom analysis data - **Agent analytics**: Filter calls by

agent_id

and analyze performance - **Success/failure analytics**: Track

call_successful

and

disconnection_reason

- **Custom alerts**: Use webhook events to t

rigger alerts based on scores, sentiment, or custom conditions [(2) ]() - **CRM integrations**: Webhook events provide real-time data for CRM syncing - **Reporting/export layer**: List Calls API supports extensive filtering with

filter_criteria

and pagination [(1) ]() ## MANAGEMENT FEATURES ⚠️ Supported: - **Agent configuration management**: Create and update agents via API (referenced in webhook documentation) [(2) ]() - **Agent version comparison**: Compare any two versions of an agent to see changes [(4) ]() Not Documented in Available Sources: - Direct API endpoints for prompt/system instruction updates - Workflow configuration management APIs - Operational settings management APIs ## Key Implementation Notes Webhook Setup for Real-Time Features: [(2) ]()

javascript
{
  "event": "transcript_updated",
  "call": {
    "transcript_with_tool_calls": [...],
    // Full call data plus transcript
  }
}

Analytics Dashboard Features: [(3) ]() - Create custom charts with column, bar, donut, line, and number visualizations - Global date range and agent filters - Integration with custom post-call analysis metrics - Real-time data updates Realistic Assessment: - 80-90% feasible for data retrieval and custom analytics - 60-70% feasible for real-time operational features (limited by phone call live monitoring) - 50-60% feasible for management features (depends on undocumented API capabilities) The built-in Analytics dashboard already provides many operational features, so focus your custom development on CRM integrations, custom alerts via webhooks [(2) ]() , and specialized analytics using the List Calls API [(1) ]() .

​ To learn more, see the sources I considered: [[1] List Calls](https://docs.retellai.com/api-references/list-calls) [[2] Webhook Overview](https://docs.retellai.com/features/webhook-overview) [[3] Get analytics insight](https://docs.retellai.com/features/analytics-dashboard) [[4] Compare agent versions](https://docs.retellai.com/agent/version-comparison) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.