Luis Garduno

ConferOS

Cloud Native Hackathon/Nov 2026

Conferences are built for people who can hear every word, understand the speaker’s language and never step out of the room. Everyone else falls behind, and there’s rarely a way to catch up. Deaf and hard-of-hearing attendees often have no live captions, or they have to arrange access ahead of time. Non-native speakers lose the thread in fast talks. Anyone who takes a call, arrives late or switches sessions misses what was said, and once it’s been said, it’s gone. Important updates like schedule changes, room locations or submission deadlines are announced once from the stage, so people end up asking the same questions over and over. Organizers answer those questions by hand, and speakers have no easy way to make their talk accessible. ConferenceOS turns what is said in each room into a live, shared record. Attendees get live captions on their own phone or laptop, on-device translation, and a searchable transcript. They can ask “when is lunch?” and get an answer taken from what was actually said, with a link to the moment it was said. An experimental signing avatar explores how sign output could work, though it is not yet a replacement for a human interpreter. Nobody has to request access in advance or be in the room at the right second to know what’s going on.

Ctrl-Alt-Build·@mardecopas,@minhoru123,@rmirza