From Research to Cutting-Edge Technology!
Ekkono is a Swedish software company founded in 2016 by Rikard König, a PhD and Senior Lecturer in Machine Learning, and serial entrepreneur Jon Lindén, together with Anders Alneng and Joakim Andersson, both previous co-workers to Jon. Ekkono is a diverse team of experienced entrepreneurs, Machine Learning Engineers, Data Scientists and Product Managers – on a mission to use technology to do good.
Ekkono is born out of seven years of machine learning research, specifically high performance computing research and predictive analytics. The outcome is a resource efficient small-footprint solution that can run most applicable machine learning techniques on small platforms, i.e. an edge computing platform for connected things – IoT – packaged in a Software Development Kit.
On the Edge
Ekkono’s edge machine learning software – that runs onboard the machines, vehicles and other things – enables smart, self-learning and predictive features. While the traditional machine learning approach is to collect big data from many units over a long period of time to find common denominators, we turn it around. We learn what’s normal for an individual device, we see when something deviates from that normal, and we learn more and more over time to make the machine learning model adaptive and more precise.
Our toolkit facilitates use cases like:
- Predictive/Condition-based maintenance
- Self-configuring products
- Performance optimization
- Entirely new business models
With Ekkono you can harmonize on one solution for all your edge machine learning needs.
The Internet of Things is a genuine transformation of how product companies do business.
Consumption-based pricing, added-value services based on domain expertise, redefining the reseller chain, and more. But this can’t be done manually. It doesn’t scale. It requires automation.
The solution for IoT automation spells machine learning. The traditional approach to machine learning, however, where data is collected and processed in gigantic cloud-based data lakes, comes with too many problems and constraints. Ekkono solves this problem by using machine learning in a different way. Instead of looking for common denominators from many, we learn what’s normal for the individual device by running onboard the device – edge machine learning. And we really do edge machine learning, not just edge inference. Our core business is about predictions, and this is happening now.