Skip to main content
This guide covers call recording, transcription, conference calling, speech recognition, text-to-speech, answering machine detection, audio streaming, and number masking.

​
Call Recording

​
Record Outbound Calls

Use the record parameter in the Make Call API.

​
Record Inbound Calls

Use the <Record> XML element in your answer URL response.

​
Record Both Legs

For calls bridged with <Dial>, use the record attribute on the Dial element.

​
Recording Storage

  • Recordings stored on Plivo servers
  • Access via API or console
  • Download URLs provided in callbacks

​
Delete Recordings

Use the Recording API with the recording ID: 
DELETE /v1/Account/{auth_id}/Recording/{recording_id}/

​
Transcription

Plivo can automatically transcribe recordings.
FeatureSupport
LanguagesEnglish only
Other languagesNot transcribed
Blank transcriptionMay occur for short recordings (<19 seconds)

​
Text-to-Speech (TTS)

Convert text to spoken audio using the <Speak> XML element.

​
Supported Features

  • Multiple languages and voices
  • SSML markup for pronunciation control
  • Speed and pitch adjustments

​
Usage

<Response>
  <Speak>Hello, welcome to our service.</Speak>
</Response>
See Speak XML Reference for voice options.

​
Speech-to-Text (STT) / ASR

Automatic Speech Recognition transcribes caller speech in real-time.

​
Use Cases

  • IVR systems with voice commands
  • Voice search
  • Transcription during calls

​
Supported Languages

Check ASR documentation for current language support.

​
Integration

Configure ASR in your application’s XML response or via streaming.

​
Answering Machine Detection (AMD)

Detect whether a call is answered by a human or machine (voicemail, IVR).

​
Detection Types

ResultMeaning
HumanLive person answered
MachineVoicemail or automated system
UnknownDetection inconclusive

​
Configuration

Enable AMD in the Make Call API with:
  • machine_detection: Enable/disable
  • machine_detection_time: Detection window (ms)
  • machine_detection_url: Callback URL for results

​
Best Practices

  • Use async mode to avoid call delays
  • Handle “unknown” results gracefully
  • Test with your specific call patterns

​
Conference Calling

Create multi-party audio conferences.

​
Create a Conference

Use the <Conference> XML element: 
<Response>
  <Conference>MyConference</Conference>
</Response>

​
Conference Features

FeatureDescription
Mute/UnmuteControl participant audio
HoldPlace participants on hold
RecordingRecord entire conference
CallbacksEvents for join, leave, speak

​
Manage Conferences

Use the Conference API to:
  • List active conferences
  • Kick participants
  • Mute/unmute members
  • Play audio to conference

​
Audio Streaming

Stream real-time call audio to external services.

​
Use Cases

  • Live transcription services
  • Real-time analytics
  • AI/ML processing
  • Custom recording solutions

​
Setup

Use the <Stream> XML element to connect to a WebSocket endpoint: 
<Response>
  <Stream bidirectional="true">wss://your-server.com/audio</Stream>
</Response>

​
Stream Data

  • Audio format: PCM (configurable)
  • Bidirectional streaming supported
  • Metadata included in stream
See Audio Stream documentation for integration details.

​
Number Masking (Sessions)

Connect two parties without revealing their phone numbers.

​
How It Works

  1. Create a session with two parties (Party A and Party B)
  2. Plivo assigns a virtual number
  3. When Party A calls the virtual number, it connects to Party B
  4. Neither party sees the other’s real number

​
Use Cases

  • Ride-sharing (driver ↔ passenger)
  • Delivery services (courier ↔ customer)
  • Marketplaces (buyer ↔ seller)
  • Dating apps

​
Session Management

Create and manage sessions via the Sessions API.

​
Key Parameters

ParameterDescription
virtual_numberPlivo number used as proxy
first_partyPhone number of Party A
second_partyPhone number of Party B
session_expirySession duration
virtual_number_cooloff_periodTime before number can be reused
force_pin_authenticationRequire PIN to connect

​
Session with Single Party

Create sessions where only one party is defined initially—the second party can be added later.

​
Update Party Details

Modify party information in active sessions via API.

​
Costs

Replacing sub-optimal numbers or extending sessions may incur additional charges.

​
Voice Alerts

Monitor call traffic and application health.

​
Alert Types

AlertTrigger
Call failure spikeUnusual increase in failed calls
Invalid XMLApplication returning malformed XML
High latencyResponse time exceeds threshold
Traffic anomalyUnusual call patterns

​
Configure Alerts

  1. Navigate to Voice > Settings > Alerts
  2. Enable alert types
  3. Configure thresholds
  4. Set notification preferences

​
DTMF

Dual-Tone Multi-Frequency tones for IVR navigation.

​
Capture DTMF Input

Use the <GetDigits> XML element: 
<Response>
  <GetDigits action="/handle-input" timeout="10">
    <Speak>Press 1 for sales, 2 for support.</Speak>
  </GetDigits>
</Response>

​
Send DTMF

Use the DTMF API to send tones during a call.

​
Limitations

  • DTMF not supported in China
  • Supported format: RFC-2833