Serverless WebRTC for Voice AI 🗣️ - `SmallWebRTCTransport` in @pipecat.bsky.social 0.0.63
We wanted to make it easier to use WebRTC while you're prototyping, or for small-scale, experimental, or customized deployments.
Using Gemini search metadata in a voice AI application
Filipi added support in Pipecat for Google Gemini's `groundingMetadata`. This makes it easy to do things like:
- link to URLs
- log searches for observability
- use specific search result chunks for RAG
youtu.be/oL9w-3Hbag0
Developer meetup on Tuesday for Voice AI folks in SF. Strong panel - wish I could be there in person @kwindla.bsky.social @dailyco.bsky.social.
For anyone building with Pipecat AI, this might be worth going along to ᓚᘏᗢ
Every time I pair with Aleix, I marvel at his Emacs skills and feel both motivated and inadequate 😆
Exciting launches for conversational AI: Gemini 2.0 and new voice-to-voice + video input API.
From Pipecat:
➡️ New client libraries for Web, React, Android, iOS, C++.
➡️ Pipecat service for Gemini Multimodal Live API.
➡️ Snazzy Gemini starter kit: github.com/pipecat-ai/g...
@langchain.bsky.social would love to work more on the LangGraph support in Pipecat too.
A community contributor built a framework implementation, which is has a lot of use-cases attached, and we're seeing it used more and more.
Anyone I can connect with about this?
🚀 Pipecat Flows landed this week!
Mark and the team would love some feedback. We’re thinking a lot about:
➡️ Intuitive, user-friendly ways to build complex realtime voice AI workflows.
➡️ Ways to create structured, multi-turn conversations for your Pipecat voice agents.
Yeyey hackathon week for Daily Bots. AI-powered codec from Metal Gear Solid? There's a bot for that.
🔉 Watch with sound
➡️ Try it out mgsx-daily-bots.vercel.app
➡️ Code: github.com/jptaylor/mgs...
A few bugs still to squash, but try it out!
