Author Archives:

Announcing Genkit Middleware: Intercept, extend, and harden your agentic apps

Genkit is an open-source framework designed to help developers build production-ready, agentic AI applications using TypeScript, Go, Dart, and Python. The framework utilizes a powerful middleware system that intercepts generation calls to inject custom behaviors like retries, model fallbacks, and human-in-the-loop tool approvals. By attaching hooks at the generate, model, and tool layers, developers can ensure high reliability and deterministic control over model outputs. Furthermore, Genkit allows for the creation and stacking of custom middleware, all of which can be inspected and debugged through a dedicated Developer UI.
This entry was posted in Uncategorized on by .

Bring Native Visibility to Your VoIP App Experience with Telecom’s Latest Alpha

Posted by Nataraj KR, Android Developer Relations Engineer


The initial launch of the Jetpack Telecom library introduced CallsManager, replacing the legacy ConnectionService API to simplify VoIP integration. CallsManager streamlines call lifecycle management and audio routing while enabling interactions with remote surfaces like smartwatches, Bluetooth devices, and Android Auto. Additionally, it supports call extensions for richer features—such as participant handling, custom icons, call silencing and meeting summary on remote surfaces—all while maintaining backward compatibility down to Android O (API Level 26).

Building on this foundation, Jetpack Telecom v1.1.0 brings native-level visibility and convenience to third-party VoIP apps. This latest release introduces powerful new capabilities, including unified call history, call log exclusion, and native callback functionality, making call management more seamless than ever for users.

Here is a closer look at what’s new and how you can implement these features in your applications.

Bridging the Dialer Gap: Unified Call History and Callbacks

Historically, users have had to open individual third-party apps to view their VoIP call history or return a missed call. With the new integrated call logging feature, system dialer apps can now surface call logs directly from third-party VoIP apps.

Even better, users can now initiate a callback to a VoIP contact straight from their native system dialer, streamlining the communication experience.

How it works:

To opt-in to this feature, do the following:

  • Register for Callbacks: Your VoIP app must register a new system-protected intent: TelecomManager.ACTION_CALL_BACK.
  • Log the Call: Use TelecomManager.addCall (or related Jetpack APIs) to ensure the system automatically logs the call.
  • Manage Call IDs: When a call is registered, CallControlScope.getCallId provides a unique UUID. The system dialer uses this exact TelecomManager.EXTRA_UUID when creating the callback intent.
  • Initiate the Callback: Your application must store and manage the call details associated with this UUID. When the system dialer fires the callback intent with the EXTRA_UUID, your app can seamlessly resolve the ID and initiate the call with the correct details.


Fine-Grained Control: Call Log Exclusion

We recognize that not every VoIP call should be visible in the system's native dialer history. Whether for privacy reasons, ephemeral communication, or app-specific behavior, you need control over what gets surfaced.

To address this, we are introducing Call Log Exclusion. You can now prevent specific calls from being logged into the system call logs by setting the isLogExcluded boolean to true within CallAttributesCompat. By configuring this flag, the call remains completely hidden from the system logs, and the native dialer will not display it.

Important Note on Compatibility
These integrated logging and callback features are available for devices running Android 16.1 (SDK 36.1) and higher. Refer here to compile your app with Android SDK 36.1.

Get Started

We encourage developers to test these integrations and explore how unified call history and callbacks can improve the daily user experience of your VoIP applications.

To help you get started and see these APIs in action, we have put together a sample application demonstrating the new integrations.

View the sample app here: https://github.com/android/platform-samples/tree/main/samples/connectivity/telecom

Check out the release notes and documentation to start implementing these features today!

Note: Although Jetpack Telecom v1.1.0 APIs are accessible for integration, the system dialer's ability to render native call logs is being introduced in phases, beginning with Google Meet. To safeguard against spam, native dialers utilize secure package allowlists to control VoIP display. For local testing of your callback and logging implementations, we recommend using the open-source Telecom Sample Dialer app as your emulator environment.

This entry was posted in Uncategorized on by .

Expanding language support for refining messages with Gemini in Google Chat

We are expanding the capabilities of Gemini in Google Chat by adding support for several new languages when refining message drafts. In addition to English, users can now use Gemini to polish their messages in French, German, Italian, Japanese, Korean, Portuguese, and Spanish. This update allows a broader range of global users to improve the clarity and professional tone of their communications within Chat.

By leveraging Gemini to adjust wording, grammar, and spelling, users can communicate more effectively across different languages and regions. This is particularly helpful for teams working in multilingual environments or for individuals composing messages in a second language, ensuring that the intended meaning is conveyed accurately and confidently.

Getting started

Rollout pace

Availability

  • Business: Business Standard and Plus
  • Enterprise: Enterprise Standard and Plus
  • Consumer: Google AI Pro and Ultra
  • Other Editions: Frontline Plus
  • Education Add-ons: Google AI Pro for Education

Resources

This entry was posted in Uncategorized on by .

Accelerating on-device AI: A look at Arm and Google AI Edge optimization

Integration of Arm Scalable Matrix Extension 2 (SME2) and the Google AI Edge software stack enables high-performance, on-device generative AI by turning the CPU into a powerful matrix-compute accelerator. Using Stability AI’s "stable-audio-open-small" model as a case study, it outlines a streamlined "Convert, Optimize, and Deploy" pipeline that utilizes LiteRT, XNNPACK, and KleidiAI to automate hardware acceleration. The resulting implementation achieves over a 2x speedup in audio generation and a 4x reduction in memory usage while maintaining high audio quality on Arm-powered mobile devices and laptops.
This entry was posted in Uncategorized on by .

Expanded options for header images in Google Forms

Form creators now have more options for header images in Google Forms. Previously, creators could only use a pre-selected theme, upload an image, or choose an image from Google Photos. In addition to these options, they can now choose an image from Google Drive or Google Images or use a URL. This makes it easier to add an appropriate and compelling header image to your form.



User chooses a Google Drive image as the header image for a Google Form

Getting started

Rollout pace

Availability

  • Available to all Google Workspace customers, Workspace Individual subscribers, and users with personal Google accounts

Resources

This entry was posted in Uncategorized on by .

Improving security posture with default context-aware access for all SAML applications

Administrators can now apply a global context-aware access (CAA) policy to all SAML applications within their organization. This update introduces a default assignment that serves as a universal security baseline, automatically protecting any SAML-based app that does not have a specific policy already assigned. By establishing this "secure-by-default" posture, IT teams can help protect internal data and third-party SaaS tools as new applications are integrated into their ecosystem.

This global control significantly reduces the administrative burden of managing security for applications at scale. Instead of manually configuring rules for every individual SAML app, administrators can set a single policy to cover their entire environment. Specific application-level policies will still take precedence, allowing for granular control where needed while the global policy acts as a reliable safety net.

These default policies support both Monitor and Active modes, providing flexibility in how security requirements are phased in. Detailed audit logs will capture these enforcement events, and remediation messages help end users understand how to resolve access issues independently.

Admins can configure CAA policies for all SAML apps in the Admin console under Security > Context-aware Access > General settings

Admins can configure CAA policies for all SAML apps in the Admin console under Security > Context-aware Access > General settings.

Getting started

Rollout pace

Availability

  • Enterprise: Enterprise Standard and Plus
  • Education: Education Standard and Plus
  • Other Editions: Frontline Standard and Plus; Enterprise Essentials Plus; Cloud Identity Premium

Resources

This entry was posted in Uncategorized on by .

Enhancing the Datadog app for Google Chat

The latest version of the Datadog app for Google Chat allows you to integrate Datadog into your Google Chat workflows more seamlessly and stay on top of critical infrastructure without leaving the flow of team collaboration. Newly added features include: 

Link previews
Cut and paste a widget from a Datadog dashboard and see the chart inside Google Chat.

Image of link preview generated by Datadog



Improved notification set-up
Link your Workspace domain to Datadog and you’ll be able to easily create Google Chat notification handles from within Datadog.

Image of Datadog setup and notification



Incident management
Declare incidents with a command from Google Chat, and automatically create spaces in Google Chat to centralize conversations about an incident.

GIF showing example of incident management with Datadog in Chat


Note that a Datadog license is required to use the Datadog app for Chat.

Getting started

Rollout pace

Availability

  • Available to all Google Workspace customers, Workspace Individual subscribers, and users with personal Google accounts (Datadog license required)

Resources

This entry was posted in Uncategorized on by .

Chrome Dev for Desktop Update

The Dev channel has been updated to 150.0.7838.0 for Windows, Mac and Linux.

A partial list of changes is available in the Git log. Interested in switching release channels? Find out how. If you find a new issue, please let us know by filing a bug. The community help forum is also a great place to reach out for help or learn about common issues.

Chrome Release Team
Google Chrome

This entry was posted in Uncategorized on by .

Announcing expanded experiment functionality in Google Ads API v24.1

We are excited to announce a significant expansion of the Google Ads A/B experiments feature set in the v24.1 release, launched yesterday, May 13, 2026. We know API functionality is critical for advertisers who manage large sophisticated accounts, and we want to give these advertisers the capability to better create and manage their A/B experiments at scale. This update broadens the platform's current scope, enabling advertisers to leverage the power of experiments for a broader range of testing scenarios to drive campaign performance.

Direct experiment reporting

One of the most requested enhancements is the ability to query experiment statistics directly through the API. Previously, you had to manually calculate statistics by querying campaign resources. With v24.1, we are exposing experiment arm-level and significance stats, including:

  • Control arm stats: The average value of metrics in the control arm.
  • Point estimates: The mean value of the difference between the control and treatment arms, calculated via the jackknife method.
  • Margin of error: The radius of the confidence interval.
  • P-values: Represents the calculated probability that the difference could have occurred by chance.

New experiment types

We are expanding support for a variety of experiment types to help you find the best configurations for your business:

  • Feature-specific tests (AI Max and broad match): Test the impact of enabling AI Max or broad match keywords. These use an intra-campaign workflow, where the feature is tested within the original campaign shared by both the control and treatment arms.
  • Custom video experiments: Run custom experiments consisting of Video campaigns.
  • Performance Max replacement: Test how your Shopping campaigns perform compared to Performance Max.
  • Performance Max asset optimization: Specifically for Performance Max, you can test different variations of assets including text, images, and videos. This asset optimization workflow helps determine which creative combinations drive the best results.

While these updates cover a significant number of use cases, we are continuing to work on the platform and plan to introduce even more experiment types in future releases.

Where can I learn more?

Updated documentation and code examples are available on the Google Ads API Developer site. We also encourage you to join the discussion and share your feedback on our Ads and Measurement Community Discord server.

 - Laura Chevalier, Google Ads API Team
This entry was posted in Uncategorized on by .