Product changelog and announcements

Answering machine detection (AMD) can now be enabled when creating child calls using <Dial><Number> and <Dial><Sip>, or when creating a new conference participant via POST to the /Participants API for accounts with Enhanced Programmable SIP Features enabled.

AMD will run in the background while the call connects and will return the results of our detection to inform your application when a human or machine answered, or when the voicemail message beep has been detected.

Answering machine detection is $0.0075 per completed call.

We are excited to announce that we have released a Go Helper for every Twilio API built using OpenAPI.

Twilio's Go Helper Library major version 1.0.0 is now available.

To learn more about check out our getting started guide.

Twilio SendGrid released Inbound Parse API 2.0 on October 6, 2022. Inbound Parse 2.0 modernizes the Inbound Parse API’s codebase, enabling us to quickly make feature updates and fixes. This release does not impact API usage or integrations, and no customer action is required. Twilio SendGrid will migrate customers to Inbound Parse API 2.0 in the coming months. We will closely monitor the migration to protect your email program from any unexpected impacts.

A2P 10DLC Messaging customers with a Tax ID (for example, EIN in the US) can now register a Low Volume Standard Brand in Twilio Console - Trust Hub - A2P Messaging

Twilio has deprecated the ability to register new Starter Brands due to updated registration requirements from The Campaign Registry (TCR). As a result, the option to create or update Starter Profiles or Brands is disabled in the Twilio Console and APIs. API requests to create or update Starter Profiles will fail with error code 22404; API requests to create or update Starter Brands will fail with error code 21726.

Previously registered Starter Brands and Campaigns are subject to new data validation requirements. Any campaigns and brands that do not meet those requirements will be deleted by Twilio on behalf of customers starting Nov 3rd.

Customers with Tax IDs are requested to register for A2P 10DLC using the Low Volume Standard Brand or Standard Brands options.

For more information, please check out our recent blog post which explains the process for new registrations and the criteria for keeping existing registrations.

Reusable Tracking Plans & Libraries includes two new features: Property Library Syncing and Duplicate Tracking Plans & Libraries. Both beta features are available to all customers with access to Protocols.

Property Library Syncing enables Protocols users to apply changes made to a Property Library across any synced Tracking Plan.

Duplicate Tracking Plans enables Tracking Plan users to duplicate existing Tracking Plans and libraries.

For more information, see Property Library Syncing and Duplicate Tracking Plans & Libraries.

This month, Bento released a new integration with Segment, which is in beta as of September 21.

For more information, see the Bento Destination documentation.

For questions and support on Destinations, please reach out to the Destination partner. 

The Developer Portal is the place where partners go to manage their integrations with Segment.

Once invited, partners can log into the Developer Portal where they can view their integration’s status, manage their integration’s catalog metadata, and invite additional team members to join their team. This feature is only available for partners.

For more information, see Segment's Developer Portal.

In this version of Flex UI we have updated the Status Report, formerly known as the Debugger UI, to now include a summary of Flex UI's operation status for each user. The Status Report is now available in GA and includes debugger logs, Twilio service status information and whether Flex UI is running in degraded mode.

Find out more about the updated Status Report on the Troubleshooting Flex UI page.

We have added two new speech recognition models to <Gather>: experimental_conversations and experimental_utterances. Experimental models are designed to give access to the latest speech technology and machine learning research, and can provide higher accuracy for speech recognition over other models, but do not support all languages and lack some features like confidence scores.