July Meetup: AI Song Generation and surfing attention

It's another iteration of our monthly meetup. This month, we are looking forward to a long talk by Arjun Bahuguna from Audio Realities and a short talk by Eugene Yakshin from hvoya audio. See talk details below.

As always, there will be time to network over drinks and pizza after the talk. Anyone interested in Audio development is welcome!
Attendance is limited to people who have RSVP'd. If the event is full, please register for the waiting list in case a spot opens up.

AI Song Generation Demystified: Under the Hood of a Text-to-Music System - Arjun Bahuguna
Ever wondered how systems like Udio and Stable Audio turn a text-prompt into a full-fledged song? In this talk, we’ll pull back the curtain on the technology behind text-to-music generation, focusing on latent diffusion models. We’ll compare popular model architectures, break down key concepts with intuitive visuals, and explore the “why” behind their design choices. No deep learning background needed - just curiosity! We’ll end with a short interactive quiz to recap and test your understanding.

Arjun is the co-founder of Audio Realities, an audio tech startup based in Aachen, where he leads the development of multilingual voice conversion models. He previously worked at Pitch Innovations, a plugin company based out of Chennai, where he worked on machine learning for music and plugins. He’s been exploring deep learning since 2016, and his work has been featured at NeurIPS, ADC Bristol, ADCx India, and will be presented at DCASE 2025 in Barcelona.

Surfing Attention, Playful Control - Eugene Yakshin
I spent many hours listening closely to the decaying tails of shimmer, noticing patterns that lie outside the description of individual components. It turned out, new properties of sound aren't always found in the algorithms, but in the way of listening. Which approaches to investigation allow us to perceive these emergent params?

Eugene Yakshin is a C++ developer and an artist, combining his interests in complex systems and music in audio programming. He worked at companies like Endel and Ultimate Guitar, creating generative music engines, audio plugins and music analysis libraries. His Hvoya Audio tools are unusual structures formed by well known elements.