# Need advice on approach for bidirectional AI streaming app

**URL:** <https://devforum.zoom.us/t/need-advice-on-approach-for-bidirectional-ai-streaming-app/131482>\
**Category:** Meetings\
**Created:** [April 14, 2025, 8:41pm UTC](https://devforum.zoom.us/t/need-advice-on-approach-for-bidirectional-ai-streaming-app/131482 "2025-04-14T20:41:27Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![Steve4m](https://sea2.discourse-cdn.com/flex016/user_avatar/devforum.zoom.us/steve4m/32/68552_2.png) [@Steve4m](https://devforum.zoom.us/u/Steve4m)\
**Post date:** [April 14, 2025, 8:41pm UTC](https://devforum.zoom.us/t/need-advice-on-approach-for-bidirectional-ai-streaming-app/131482/1 "2025-04-14T20:41:27Z")

</div>

I’m building a bidirectional, conversational, voice ai app that will participate in Zoom meetings. I tried the web meeting SDK and my bot is joining the meeting. However, it seems that I need to use the Video SDK to access the video and audio live stream. I need the raw pcm audio stream that will be sent to my GenAI backend.

The documentation on the live stream use case isn’t great and bidirectional is almost non existent. I’m not 100% sure that I can achieve a working demo of this app using the web approach versus building a native app. Please can someone advise on which SDK I can use to achieve this use case in the simplest way. Any samples or demoes would be appreciated.

---

<div class="post-metadata">

**Author:** ![chunsiong.zoom](https://sea2.discourse-cdn.com/flex016/user_avatar/devforum.zoom.us/chunsiong.zoom/32/35488_2.png) [@chunsiong.zoom](https://devforum.zoom.us/u/chunsiong.zoom)\
**Post date:** [April 16, 2025, 7:51am UTC](https://devforum.zoom.us/t/need-advice-on-approach-for-bidirectional-ai-streaming-app/131482/2 "2025-04-16T07:51:33Z")

</div>

@Steve4m , if you are building something which needs to be bidirectional, you will probably need to use Meeting SDK.

In this case, you will either need to use Windows Meeting SDK or Linux Meeting SDK to get raw PCM audio, pass it to your backend service, generate a PCM audio and use raw data to send raw audio back to the meeting.

---

<div class="post-metadata">

**Author:** ![Steve4m](https://sea2.discourse-cdn.com/flex016/user_avatar/devforum.zoom.us/steve4m/32/68552_2.png) [@Steve4m](https://devforum.zoom.us/u/Steve4m)\
**Post date:** [April 16, 2025, 9:01am UTC](https://devforum.zoom.us/t/need-advice-on-approach-for-bidirectional-ai-streaming-app/131482/3 "2025-04-16T09:01:16Z")

</div>

thanks for confirming. Does that mean that there isn’t a MacOS SDK that can handle this use case?

---

<div class="post-metadata">

**Author:** ![chunsiong.zoom](https://sea2.discourse-cdn.com/flex016/user_avatar/devforum.zoom.us/chunsiong.zoom/32/35488_2.png) [@chunsiong.zoom](https://devforum.zoom.us/u/chunsiong.zoom)\
**Post date:** [April 16, 2025, 11:40am UTC](https://devforum.zoom.us/t/need-advice-on-approach-for-bidirectional-ai-streaming-app/131482/4 "2025-04-16T11:40:44Z")

</div>

@Steve4m macOS SDK would work as well. From a hosting and deployment point of view, windows and Linux are the more common server OS
