The mission of echosonic

At echosonic, we pursue a greener, securer and more human-orientated future of machine learning. By bringing the machine learning to MEMS microphones, echosonic-backed smart microphones offload the bandwidth and reduce the power consumption for cloud-based human voice and sound event recognition. Without data leaving the microphones, we address privacy and security concerns while achieving exceptional processing speed and inference accuracy compared to legacy smart devices.

​

The on-edge sound processing is an end goal for the next generation of smart devices, but incumbent solutions are limited by the chips’ cost and infrastructure required to support the ML model complexity. echosonic innovative approach avoids the software complexity by executing most of the computation in the analog and physical domain.