Out-of-the-box deep learning (DL) models are extremely inefficient in the sense that they are too large, too slow, and consume too much power to deploy into production.
We build a software framework for developers, that makes it easy to build and tweak DL models, such that they run particularly efficient on target hardware platforms. Additionally, the developed model will be offered such that it is ripe for production and ready for deployment in a push-button fashion.