Deepgram has just secured a funding round which included prominent players like Y combinator to advance its voice ai technology. Deepgram aims to be the AI that revolutionizes the restaurant and drive-in industries.

The company’s technology distinguishes itself high accuracy across diverse accents and a level of realism that mimics human social cues. Deepgram’s models can handle choppy conversations and natural pauses, ensuring that AI interaction is fluid, rather than clunky. To achieve this, the company the company prioritizes ultra-low latency, targeting response times of under 500 milliseconds.

Deepgram wants to launch its own “Deepgram as a service” for the restaurant industry, with their technology powering retail conversations and automating drive-thru services. This is a rather complex task given that major players like McDonalds have struggled with implementing their own voice AI agents. Deepgram’s on the ground approach to difficult acoustic environments, like noisy drive throughs will bridge the gap between mass adoption and failure.

Deepgram also has a robust API suite, Aura-2 (a text to speech model), Nova-3(speech to text model) and Flux(conversational recognition). The APIs are already being used by 1300 enterprise clients.

Deepgram is building up its voice AI community by creating a physical “Voice AI Collaboration Hub” in San Francisco to host developers and industry leaders for demonstrations and hackathons. The company aims to be the cornerstone of voice AI like how Stripe and Twilio are to payments and communications respectively.

In the near future, instead of interacting with a restaurant employee for taking orders, we would be interacting with AI agents. Who knows, this might just simplify the whole ordering and payment process.