Before Sesame had an app, it had a voice. In February 2025, the company put Maya and Miles online for people to call. Ryan Brown, one of Sesame’s three founders, shared the demo with an unusually useful admission. Natural computer voices, he wrote, could open up positive experiences for many people, “but we have a long ways to go.” The statement reads like an engineer’s launch note: happy to let the public hear the work, unwilling to pretend the work was finished.
For Brown, the task was also a change of scenery. He had spent years in the world of headsets, first at Oculus and then at Meta Reality Labs. Those machines place a screen close to the eye and ask the wearer to enter a constructed space. Sesame is chasing a different daily habit. Its agents talk back, keep a thread of conversation and, eventually, are meant to accompany their users through a pair of glasses. The ambitions are large; the public evidence comes in smaller increments: a voice demo, a mobile preview, research on conversational timing, and a promised eyewear launch.
That order is worth noticing. A device you wear all day must earn a place in all the unremarkable minutes of a day. A voice assistant that cannot handle an interruption or a pause would make those minutes feel longer. Brown’s background sits precisely at that awkward meeting point between an impressive prototype and an object people actually want to keep using.
The hardware years
Brown’s path into this company ran through the earlier promise of a new computer interface. He worked at Oculus, the virtual reality company that became part of Facebook, and later directed research engineering at Meta Reality Labs. Reporting on Sesame describes him as a hardware architect at Oculus. The broad shape of the career is clear: he spent a long stretch making ambitious devices contend with the stubborn details of engineering.
Oculus was a useful apprenticeship for a new kind of wearable. A headset is not simply a display with a strap. It is a product people judge through its weight, comfort, responsiveness and the seamlessness of many parts acting together. Sesame has a different physical design problem, but the demand for integration remains. Its own team page says useful voice products require hardware, software and machine learning to work closely together. The sentence is almost austere for a company describing an ambitious future. It also explains why a hardware veteran belongs in its founding trio.
Brown’s public profile places him in Bellevue, Washington, one of Sesame’s office cities. His education entry lists Texas A&M University from 1997 to 2001. He also shared a colleague’s memory of working at Tesla in 2013, when engineers competed for the chance to take an engineering Model S home. The colleague wrote that Sesame had brought back a similar desire to use the product after work. Brown passed the post along while recruiting for hardware and operations roles. It is a telling measure for a builder: the office device has become something somebody wants for a weekend.
“Giving computers a natural voice will unlock so many positive experiences for so many people - but we have a long ways to go.”Ryan Brown, on Sesame’s first voice demo
Three founders, one conversation
Sesame says it was founded in 2023 by Brown, Brendan Iribe and Ankit Kumar. Iribe is an Oculus co-founder; Kumar came from work on augmented reality and conversational AI. The partnership joins people who know the difficulty of producing a device with people focused on how a computer speaks. Brown’s place in it is distinctive. He had already worked alongside the Oculus generation of builders, then spent time in the research engineering arm of a much larger company. Sesame gives that experience a different destination.
The team grew in public view. Nate Mitchell, another Oculus co-founder, joined as chief product officer in 2025 and described the move as getting the band back together with Iribe, Brown and other old colleagues. One can hear the reunion in that line, but it is the new assignment that matters. A headset asked how convincingly a computer could create another world. Sesame’s glasses will be judged on how gracefully a computer accompanies this one.
There is a small joke hidden in the company’s development sequence. To build eyewear, Sesame first asked people to close their eyes, metaphorically, and listen. Maya and Miles were the first introduction. The public did not need a frame, a lens or a battery specification to decide whether the voices were worth another conversation. Brown introduced them on LinkedIn as expressive AI voices that sounded natural and conversational. He also invited people working in AI, speech and elegant hardware to join the company. In the same post, a voice clip let Maya banter about the prospect of telling a joke. The demonstration showed what the company wanted the medium to feel like: responsive enough to play along.
A pause is a product decision
Voice sounds simple because human conversation is familiar. The engineering is less forgiving. An agent must know when a person has finished, when a short sound means “keep going,” and when an interruption means “stop.” It has to answer quickly enough to feel present while taking enough time to say something useful. Sesame’s researchers made that problem explicit in August 2026 with TurnBench, an open benchmark for the timing of spoken dialogue. The team wrote that the typical gap between human turns is about 200 milliseconds. That is a sliver of time in which a machine must make a social judgment.
Brown has not claimed personal authorship of that research. Its presence in Sesame’s public work nonetheless helps explain the company he is building. A wearable agent’s quality will be measured in thousands of such little moments. If it talks over a user, the hardware becomes an annoyance. If it waits too long, the conversation becomes a queue. Voice design is a question of manners as much as a question of model output.
Sesame’s May 2026 iOS preview gave those voices more to do. Maya and Miles were joined by Simone and Charlie. The app offered voice and text, search, notes, reminders and summaries after calls. The company said the preview was available in 39 countries and that an Android version was on the way. For a product meant to travel with someone, a phone is the obvious proving ground. It can test whether conversations remain useful when people are commuting, walking or trying to remember an idea, before the assistant moves onto a face.
Brown greeted that release by praising the “thought, care and effort” behind the first iOS experience, then singled out Mitchell and the product team. It is a modest kind of founder’s announcement. The company was putting a larger part of its proposition into people’s hands; Brown’s public note drew attention to the colleagues who made that step possible.
The glasses have to follow the voice
Sesame says its first eyewear collection is coming in 2027. That date is a target, not a product review. The company has described lightweight glasses intended to give people convenient access to an agent while moving through the world. The concept puts Brown’s two professional chapters in the same frame. The hardware experience of Oculus and Reality Labs meets a voice system designed around natural timing, memory and an ability to respond to the person in front of it.
It also raises the ordinary questions that decide whether a new device survives beyond its first demonstration. Can someone wear it for hours? Does the assistant help in a moment when reaching for a phone feels cumbersome? Will the voice know when to remain quiet? Those questions do not yield to a spectacular video. They yield to use, revision and the unglamorous work of making many disciplines agree with one another.
Brown’s public career gives this project a through line. He has worked on machines that change the way a person meets a computer. At Sesame, the visible steps so far are deliberately incremental. The first impression was a voice. The next was an app with useful tasks and four personalities. A research release measured the timing that makes a dialogue flow. The glasses remain ahead, with the ordinary engineering decisions still to come. For now, the company’s most revealing promise may be the same one Brown made at the voice debut: there is potential in the conversation, and there is still work to do.