For AI agents: a documentation index is available at the root level at /llms.txt and /llms-full.txt. Append /llms.txt to any URL for a page-level index, or .md for the markdown version of any page.
Login
HomeDocsAPI ReferenceMCP ServerChat SDKsRelease Notes
HomeDocsAPI ReferenceMCP ServerChat SDKsRelease Notes
  • Welcome
    • Getting started
    • Key concepts
    • Dashboard navigation
    • Agent logic
    • Improvement tactics
  • Knowledge
    • Content ingestion
    • Article management
  • Automation
    • Actions
    • Processes
    • Playbooks
    • Proactives
    • Greetings
    • Custom Instructions
    • Variables
  • Setup
    • Persona
    • Languages
    • Redactions
  • Handoffs
    • Handoff management
    • Salesforce
    • Zendesk
  • Channels
    • Chat
    • Voice
      • Voice onboarding
        • Greetings and Handoffs
        • Voice call capture options
        • Voice tools
        • Voice CSAT configuration
        • SMS consent
      • Contact center integration
    • Email
    • Social
    • Third party
  • Optimization
    • Testing
    • Conversations
    • Coaching
Login
LogoLogo
On this page
  • Overview
  • Limitations
  • Use cases
  • Capabilities & configuration
  • Capture modes for Action inputs
  • Capture options for autocapture variables in Processes
  • SMS capture consent
  • Related features
ChannelsVoiceVoice configuration

Voice call capture options

Previous

Voice AI Agent tools

Next
Built with

Overview

Voice call capture options determine how your AI Agent collects input from end users during phone calls. When configuring Actions or Processes for Voice, you can specify whether end users provide information by speaking, using their phone’s dial pad (DTMF), or responding to an SMS message.

Configuring appropriate capture modes helps set your AI Agent and your end users up for success, as different capture options suit different types of information being captured.

Limitations

Voice call capture options have the following constraints:

  • SMS requires consent: Before sending any SMS messages to an end user, your AI Agent must obtain consent. This consent only applies to sending a single SMS as part of the conversation—you cannot use it to add end users to marketing lists or send additional messages for other purposes.
  • SMS only works for mobile devices: End users may be using landline phones or may not want to receive SMS messages, so be prepared for SMS capture to not work in all cases.

Use cases

Voice call capture options support different scenarios based on the type of information being collected.

  • Collect spoken responses: Use Speech capture for text-based information like names or addresses.
  • Capture numeric input via dial pad: Use Speech and DTMF for numbers that are easier to enter than speak, such as account numbers or PINs.
  • Provide an SMS fallback: Use Speech and SMS when end users may have difficulty speaking information clearly, such as email addresses or complex alphanumeric codes.

Capabilities & configuration

Voice call capture options vary depending on where you configure them and the type of input being captured.

Capture modes for Action inputs

When configuring inputs for Actions, the available capture options depend on the input Type:

Input typeCapture options
TextSpeech: Captures input from the end user’s spoken response.
Speech and SMS: Gives the end user the option to speak the input or respond to an SMS message.
NumberSpeech: Captures input from the end user’s spoken response.
Speech and SMS: Gives the end user the option to speak the input or respond to an SMS message.
Speech and DTMF: Gives the end user the option to speak the input or use their phone’s dial pad.
BooleanSpeech: Captures input from the end user’s spoken response.

Capture options for autocapture variables in Processes

When creating autocapture variables in Processes, you can select a Voice call capture option that defines how your AI Agent collects those variable values.

Capture optionDescription
SpeechCaptures input from the end user’s spoken response.
Speech and SMSGives the end user the option to speak the input or respond to an SMS message from the AI Agent.

If your autocapture variable name reflects that the variable being captured is numerical (for example, phone number, amount), your AI Agent will also allow end users to provide this value using DTMF (their phone’s dial pad) in addition to speaking it. Before capturing a value, your AI Agent automatically tells end users that they can either say the number or enter it into their dial pad.

If you find that the AI Agent is not providing this option for an autocapture variable and you want it to, try updating your autocapture variable name to make it clearer that it is numerical.

If you find that the AI Agent is providing this option for an autocapture variable and you don’t want it to, try updating your autocapture variable name to make it clearer that it is not numerical.

SMS capture consent

Before sending any SMS messages to an end user, your AI Agent must obtain consent. Your AI Agent asks for consent before sending an SMS message.

Important considerations for SMS consent:

  • This consent only applies to your ability to send a single SMS to the end user as part of the conversation. You cannot use this consent to add the end user’s phone number to marketing lists or send them additional SMS messages for other purposes.
  • SMS only works for mobile devices. End users may be using landline phones, or may not want to receive SMS messages, so be prepared for end users not to be able to use SMS capture in all cases.

Related features

These features work with Voice call capture options:

  • Actions: Configure API integrations that collect input from end users.
  • Processes: Create workflows with autocapture variables.

Have any questions? Contact your Ada team, or email us at help@ada.cx.