AI Engineering In Audio Signal Processing

SOUND DISCOVERY: A forum for sound exploration

AI in audio moves rapidly from research prototypes to deployed real-world systems. 2026 is coined as the year of Voice AI Interaction, and Large Language Models (LLM) can now be inferred from small, embedded devices like mobile phones, hearing aids, and headphones.

However, the growing size of LLM knowledge caches are still too heavy for these platforms to support, restricting the full potential of Generative and Agentic AI to big GPU clusters.

What if we would transfer the best AI Engineering practices of LLMs to the domain of audio signal processing, without the LLMs themselves?

Join us for a free-for-members tasting session that introduces participants to the core ingredients of modern AI audio engineering, starting with Data, Models, and Algorithmic Code, and expanding toward the full lifecycle management used in production environments.

! Only 10 seats available !

Date: August 25th, 2026

Time: 13:00 – 16:00

Language: English

Location: Aalborg University – Copenhagen

Price: Free for DSC members

For you interested in:

Audio

Digital Signal Processing

Product ownership

Business management

What to expect

Participants will build a small working pipeline during the session. You will begin with audio data preparation, move through model training and inference, and conclude with testing, automation, and deployment-ready workflows. The session is designed for engineers, researchers, and developers interested in transforming experimental audio models into robust, scalable systems. By the end of the afternoon, you will understand how the foundational components of AI systems engineering evolve into monitored, versioned, and automated pipelines.

You Will Meet

Cumhur Erkut Associate Professor at Aalborg University

Cumhur is a PhD and Associate Professor at Aalborg University Copenhagen, working on research and innovation in Audio AI Engineering, spanning machine learning, digital signal processing, immersive systems, and real-time intelligent audio applications. His work integrates data, models, and software engineering into deployable AI solutions across embedded, edge, immersive and cloud environments. He has experience in designing hands-on learning pathways that translate research concepts into industrial practice through modern engineering workflows focused on reproducibility, automation, and lifecycle management.

More info on Dr. Erkut’s current research activities can be found at https://cerkut.github.io

Razvan Paisa Postdoctoral Researcher at Aalborg University

Operates at the intersection of DSP, hardware interfaces UX, and auditory research. Explores how humans perceive and interact with sound through both auditory and tactile channels. Research focuses on haptics, music perception, and spatial audio, with emphasis on improving musical experiences for cochlear implant users. Builds experimental systems from vibrotactile concert furniture to large-scale Wavefield Synthesis setups. Background spans programming, embedded dev, HCI and interaction design. Persistently curios about the exploration of sound, touch, and perception. Outside work: kiteboards, boards of other kinds, synthesizers.

The seminar will be co-hosted by Jeppe Lindegaard, Danish Sound Cluster.