What we do, summarized for tech-savvy people


increasing the functionality of ADAU1787 (Analog Devices) by building custom DSP blocks in assembly

developing custom device drivers for Zephyr e.g. on nRF5340 (Nordic Semiconductor)

implementing custom low-latency active noise control on Hifi core (Cadence Tensilica)


building a microphone from scratch to sample sound in an 150°C airstream flowing at 10m/s in large power plant

abusing Dante (Audinate) audio network for year-long multi-channel measurement setups in fully virtualized environments.

building a loudspeaker casing out of a plastic doll, try equalising that

rewriting an algorithm for the 10th time to decrease latency by 1ms, only to rewrite it again in Assembly

data science

writing Python scripts to crunch billions of time series data points into lists of lists of lists
Disclaimer for marketing: of course including Artificial Intelligence.

linking insights to physical models, to not correlate cigarette consumption with the amount of rocket launches

we love to load all of the cores on our Threadrippers (AMD) to the maximum, which is such a contrast to working on high-efficiency applications in embedded

Our Team

Meet our interdisciplinary team with backgrounds in math, physics, computer science, biology, electrical engineering and economics.


Are you interested to become a rocket scientist? See our jobs section.

Shaping sound

The concept of eliminating noise & vibrations with its negative equivalent goes as far back as June 1933, when Paul Lueg filed his patent on “silencing sound oscillations”. Yet active noise control has never found widespread application until recently. Read our case studies to learn how you can use the technology.

Using sound

Sound is stronger than light and carries an abundance of information. To date many factories and robots still are deaf. We are here to provide them with ears and use the information to understand and interact effectively.