Deepgram acquires Pipecat-maker Daily to bolster its voice AI infrastructure

The cost to build real-time voice applications with custom AI models is significantly reduced.

Deepgram, a company specializing in voice AI, has acquired Daily, the creator of the popular open-source voice and video streaming framework, Pipecat. While the financial terms of the acquisition were not disclosed, Deepgram has committed to keeping Pipecat open-source. Daily's Pipecat framework helps manage various components of the voice AI pipeline, including speech-to-text, large language models, and text-to-speech, with the goal of reducing latency in real-time communication. This acquisition is expected to enhance Deepgram's capabilities in real-time voice and video infrastructure. Both Deepgram and Daily emerged from the same Y Combinator batch in 2016. Daily had previously raised over $62 million in funding. Deepgram recently achieved unicorn status after securing $130 million in a Series C funding round.

The acquisition signals a consolidation trend in the burgeoning voice AI sector, with companies seeking to bolster their infrastructure and expand their market reach. Deepgram's move to acquire Daily and maintain Pipecat as open-source suggests a strategy focused on fostering an open ecosystem, encouraging broader adoption of its voice AI technologies. This approach benefits developers by providing them with flexible, integrated tools, potentially accelerating innovation in AI-native software and voice agent development.

By integrating Daily's real-time communication expertise and launching the Real-Time Context Standard (RCS) protocol, Deepgram aims to standardize how voice agents interpret and utilize contextual metadata. This could lead to more expressive and context-aware voice AI applications, impacting how businesses interact with customers and how users engage with AI assistants. The emphasis on open protocols and vendor neutrality within an increasingly competitive landscape highlights a strategic bet on community-driven development and interoperability.