About Sesame
Sesame AI was founded in June 2023 by Brendan Iribe, a co-founder of the virtual reality company Oculus that was later acquired by Facebook, together with Ankit Kumar and Ryan Brown, pursuing a research agenda focused specifically on making AI voice interaction feel genuinely natural and emotionally engaging rather than the more clearly synthetic, somewhat robotic-sounding voice assistants that have been common across most prior voice AI products, based on the observation that even as large language models' text-based conversational capabilities had advanced dramatically, the actual voice output of most AI assistants still sounded noticeably artificial, lacking the natural pauses, tonal variation, laughter, and emotional expressiveness that make human conversation feel genuinely alive and engaging rather than clearly mechanical.
The company's core technology, called the Conversational Speech Model (CSM), processes text and audio information together rather than through the more traditional pipeline of first converting speech to text, processing that text with a language model, and then converting the resulting text response back into speech through a separate text-to-speech system, an integrated approach that lets Sesame's voice models, named Maya (female voice) and Miles (male voice), respond with response times of roughly 200 to 300 milliseconds while naturally incorporating conversational elements such as mid-sentence tone shifts, laughter, interruptions, and emotionally responsive intonation that adapts to the conversation's content and emotional tenor, capabilities that have made interacting with Sesame's voice companions feel notably more humanlike than most prior voice AI products, drawing significant viral attention and media coverage describing the experience as surprisingly natural and, for some users, even slightly unsettling in how convincingly human it can feel. Sesame has released several sizes of its underlying speech model as open-source software under an Apache license, while also developing plans for a lightweight wearable glasses product intended to let users access their AI voice companion throughout the day with integrated audio and eventually visual awareness of the wearer's surroundings. The company has raised a Series B funding round with participation from Sequoia Capital, following strong early user engagement that generated millions of minutes of conversation with its Maya and Miles demos shortly after their public release.
