Designed for a wide range of products
From the outset we designed the ai3™ software platform to be independent of hardware and consume as little power, memory and computational resource as possible. This inherent flexibility gives our customers the ability to explore different embedded architecture options, specific to their product ranges and strategies.
The result is incredibly powerful software that can scale and flex with our customers.
Flexible
ai3™ is hardware agnostic and requires minimal memory and computational processing, which means that it can be deployed across multiple products with differing architectures and constraints, making it suitable for a wide range of products and applications.
Scalable
The ai3™ software platform supports multiple sound profiles and enables new sound profiles to be delivered OTA. This means that consumers can enjoy new features as they are released and existing connected consumer products can be enhanced in the field for differentiation.
Accurate
Our sound recognition technology isn’t indiscriminate (like level detection) because our core technology is based on modelling the hundreds of ideophonic building blocks of sounds, rather than just dB levels. This is why our software is highly accurate and not irritating to consumers.
Proven technology
We work with the world’s most innovative consumer tech brands, who all see accurate and precise sound recognition as a critical component of artificial intelligence.

Window glass break
Dog bark
Smoke/CO alarm
Baby cry
Anomaly detection
Speech detection
Snore
Car alarm
Cough
Laugh
Car horn
Emergency vehicle siren
Doorbell
Bicycle bell
Vehicle reversing alert
Yawn
Shout
Door knock
Intruder alarm
Sneeze