AI the Mobilaris way


AI (Artificial Intelligence) is a term you see in many places today and its use span a very large range of topics. It’s definition is a slightly moving target (at least if you look for a specific, well-defined version) but the literature typically agrees to three, distinct properties of an AI. It should have the capability to learn, reason and self correct.

There are examples of AI use in products today deployed in the market, but they represent a very small share of software based systems in use. My personal view, which I believe I share with many, is that AI technologies are not just evolutionary, but revolutionary. This is not due to technologies themselves but rather the deep impact they will have on your daily life, the way we communicate (with whom are we communicating?) and the way we do business.

Don’t get me wrong, the technology development is truly fascinating and it is easy to point out key technology catalysts that has enabled the machine learning, and later AI, revolution we are seeing today. But the true measure of new technology is how it is impact the world around us. A crucial task for me as a CTO in a company which builds products very much centered around extracting value out of data, is to translate fast moving revolutionary technologies into commercial offerings that generates value to the end-user. But in doing that, it is not enough just to prove the technology and show its promise in some lab test, it needs to be deployable and manageable in operation as well.

And here lies a big challenge with AI in my view. At the centre of every AI solution, there is a model design which outlines the way the problem is addressed. That model is fed with different data points and if all goes well, it gives you an answer to the problem at hand.

However, no AI model is “ready” when it leaves the factory. It needs to be constantly updated with more learnings, learnings from its operational environment, and if you have a large number of products out on a market, the amount of collective learnings you can do is staggering. This leads to a new relationship between the customer and the vendor, where the very essence of how the software works evolves over time. This is already well known in cloud solutions today where many customers are sharing a joint software installation, but when you ship on premise software, this represents a new way of doing business.

For Mobilaris, this makes it a dual challenge. First, truly master AI technologies and recognize what kind of value they can enable for our customers. And second, make sure that you have a customer relationship and a business model that works in an AI friendly way. We are already today working with this and will launch new products within the near future which builds on AI technology, and we are engaging closely with our customers to make sure that our new products directly will generate new value.

Daniel Enström
CTO at Mobilaris



Here are some aspects around AI and Mobilaris products from a talk (20 minutes into the movie) I gave just before summer. It is in Swedish so for those of you who hasn’t yet discovered the beauty of the Swedish language perhaps you can apply some AI enabled translator to hear it in your language!