The AI Runtime - Boston
Build sound into your next AI experience.
Sound drives emotional experiences. AI can generate images, video, interfaces, and entire applications, but sound is often treated as an afterthought. Creators still have to accept whatever audio comes with a video generation or search through libraries, sync effects to the edit, and rebuild the sound manually.
For this build night, we're partnering with Mirelo AI to explore what happens when sound becomes another programmable layer of an application.
Mirelo is a European frontier research lab building state-of-the-art generative video-to-sound effect models the next audio layer for visuals. Founded by two musicians turned AI researchers, Mirelo uses the video as prompt and directly synchronizes tailored, edit-ready sounds to your footage. Backed by a16z and Index Ventures, Mirelo is redefining how audio is created for the visual world.
We'll start with a short introduction to generative audio and Mirelo, then spend most of the evening building.
You can experiment with ideas like:
Turn a silent or AI-generated video into a fully sound-designed experience
Generate synchronized foley, ambience, and sound effects directly from video
Find the right sound without searching. Best suited for commercials, launch videos, social ads, AI stories, short films and series.
Iterate on the sound without changing the visual.
Create multiple versions from one visual
Keep the sound coherent across a sequence
Build an application that generates audio dynamically
Add generative sound to games or interactive experiences
Experiment with Mirelo's API or MCP from your AI coding workflow
Combine video-generation models with Mirelo to create an end-to-end multimodal pipeline
The goal: come with an idea, build something, and leave with a working prototype.
You don't need to be an audio engineer. Developers, AI engineers, designers, creative technologists, and curious builders are all welcome.
Bring your laptop. We'll provide the technology, examples, people to build with, and plenty of room to experiment.
Sound on. Let's build.