Keen Research develops SDKs and software tools for on-device speech recognition on mobile devices and custom hardware platforms. KeenASR SDK is based on a state-of-the-art speech recognition engine with Deep Neural Network at its core, and it includes a variety of acoustic models. With our solution, all the processing happens on the device ("offline"); there are no privacy concerns and no need for backend scaling.

The SDK provides developers with a lot of control and flexibility, providing support for custom vocabulary domains, robust recognition in noisy environments, and many other features. It is currently available for iOS (Objective C, Swift), Android, and Linux.

KeenASR SDK powers a number of enterprise mobile/AR apps covering use cases such as voice-picking, augmented intelligence, checklists, workflows, interactive videos, training, etc.