Real-time generative AI

Decohere is building world-class product and infrastructure for real-time generative AI, with an emphasis on AI video. We enable our customers to iterate towards their creative vision without waiting for long-running GPU jobs.

Team Size:3
Location:Seattle, WA

Active Founders

Rishi Bhuta

Decoherence, CEO. Enabling creators to generate video from prompts with a simple, intuitive interface.

Rishi Bhuta
Rishi Bhuta

Will Stith

Co-founder and CTO of Decoherence. Working on realtime generative AI, with an emphasis on AI video. Previously at Amazon doing greenfield R&D projects in Robotics and Project Kuiper.

Will Stith
Will Stith

Company Launches

tl;dr: ✍️→🎞️

Decoherence is the easiest way to create videos with AI by using text prompts and music. We enable creators to quickly generate music videos or B-roll that matches their content.

😓 Problem

Content creators often have audio content (music, podcast, etc.) that struggles with engagement on video platforms like Youtube, Tiktok, and Instagram. They want to use AI video generation to create a compelling visual pair for their content, but existing solutions are not accessible. A combination of costly hardware requirements, programming knowledge, and constant research upkeep prevents them from using the technology.

💡 Solution

Decoherence is an intuitive, no-setup application to create all of your generative videos. It gives creators the ability to work at 10x the speed: quickly generating B-roll content or animating their music in minutes rather than days. It works by analyzing structure in generated images, converting audio signals into motion effects, applying them over time, and stitching images together to make a full video.

See the full product demo video:

🤖→🎨 Why us?

We are engineers with experience bringing bleeding-edge technology, autonomous vehicles and next-gen internet satellites, into production at a global scale.

We met at Amazon Robotics where we shared a deep appreciation for robotics because it was software that literally moves. In early 2022, we found a similar spark when we typed in a few words and received back beautiful images from AI image generators. We think things are always more fun when they move, so we applied our background in computer vision to bring the images to life in the form of video.

We learned to turn tech demos and research into products used by thousands at Amazon. We’re applying that experience to Decoherence by building a product that anybody can pick up and use.

🙌 Ask

Interested in creating video? Shoot us an email at
getaccess@decoherence.co, and mention “YC Launch”! We will bump you up the waitlist!

We are looking to chat with content creators! If you know any that may be interested in a tool like this, we would appreciate an introduction. Please reach out to rishi@decoherence.co.