Interview with Mikael Nordenstjerna, MainlyAI’s new CEO.

1. What are you looking forward to the most when starting at MainlyAI?

To get my hands dirty. I want to be in this amazing field of work and build a world-class company. With the foundation already in place, it will be a thrill to get busy realising all the promise of the tech in the company of wildly competent and inspiring people.

AI will change the world far more than most realise. I think that AI is right up there with the greatest inventions and world changing developments, such as the wheel, agriculture, writing and the internet. Joining MainlyAI will get me in the thick of things. And not only being in this exciting field, but also a chance to lead it. I strongly believe MainlyAI has a real chance of being a driver and an enabler for true change.

2. What is the impact of AI? Why is it so important to use AI in your business?

The use of AI is already widespread. You are using AI, or AI enabled, solutions and services all the time, perhaps without realising it. When you unlock your phone, the facial recognition is a product of AI. Navigation is another service that has been greatly enhanced thanks to AI. And autocorrect, although it can create involuntary humour or discomfort, is getting better and better at figuring out what you are really trying to write. Then search results, marketing, social media content, process improvements and of course chatbots are all other aspects of how AI is affecting your life today.

As a business, if you are not leveraging AI in your way of work, you risk falling behind the competition. There are so many applications where AI will be key to stay competitive. If you do not use your data and make that data into insights, you are missing the mark. Let’s list a few examples:

  • Data analysis is not limited to how many analysts you have and how good they are at their job. AI can detect patterns and deliver insights 24/7.
  • Outbound email campaigns analysis can be automated, and responses can be categorised by not only the response in itself, but tonality and other factors that a human would have to spend hours on.
  • Inventory management can be made more effective and productive by using AI image analysis. AI can detect defects, categorise, and label the inventory. Doing this manually is a cumbersome and time-consuming task.
  • AI in accounting is another field where costly errors can be avoided.

There are countless other applications of AI where the technology is a huge help. AI can be your best friend, working 24/7 with an unrivalled quality. To not use that should not be an option!

3. What do you think about the pace of AI deployment in the industry and why do you think it’s slow?

AI is on most companies’ horizon now. I think there is a fear of complexity, of how expensive it is and that they just don’t know where to start. The reality is that it does not have to be complex or expensive. I also think that some companies might not see the need. That they are doing fine without it. And that may be true today, but it will surely not be so in the future. There will be no businesses around that do not use in some way. It was the same when computers came on the scene. It was said that it was only big corporations that would have a need for computers, but time proved them wrong.

One of the main barriers for companies to start using AI is complexity. It is an abstract concept for many and daunting to even start a discussion about AI. However, today there are lots of tools and expertise out there. You don’t have to build it all from scratch. Look at others who have done successful implementations or applications of AI and learn from them. Find a way to make AI make sense to you and your company, for when you do understand its applications, it does not seem that complicated anymore.

Another barrier is the cost. The cost is as always what you make of it. You can spend millions and get nowhere, or you can start small, learn, and then grow. And as I wrote above, learn and re-use what others have done.

A third barrier is to know where to start. My advice is to look for where your company has the most obvious gain and most clear-cut case for where AI could help and start there. Once you are up and running with AI, it will be clear as to what possibilities there are, and where to apply it next. The next step is not that overwhelming anymore.

4. Where do you see companies could benefit from AI in their work to set and achieve sustainability targets?

Sustainability is something we all must have top of mind. AI can help in making companies more efficient and help them achieve their sustainable goals. It can also help with identifying which sustainability targets are achievable, which gives the best result and how they impact each other. AI can be used to minimize resources, analyse behaviours and data to recommend improvements and stop wasteful processes. This would create a gain in productivity, i.e., reduce carbon emissions for a comparable output. AI can also be used for analysing the sustainability itself and monitor, warn and suggest improvement and fixes to continuously decrease negative sustainability impact.

MainlyAI’s vision is also that we can bring down the carbon footprint of using AI itself. The emissions from running AI can be really high, but we think that a change is needed. AI generated insights, trained models, and data should be viewed as a commodity like any other and companies should be able to trade and share. This would mean that we can re-use AI commodities and thus reduce the carbon footprint. Companies should also be able to share platforms in a secure and safe way so that they are sustainable in their use of AI as well.

5. Companies are not operating in a vacuum; many processes are similar. What would be the natural next step in development of AI?

Every company likes to think of themselves as unique. I hate to break it to you, but most companies work in similar ways, and have the same challenges and problems. AI is no different from any other commodity a company uses. To mitigate many of the challenges with AI, the next step would be to be able to share and trade insights, data and algorithms in a safe and secure way as described above. If AI models are possible to re-use, there is no new training that needs to be done. The carbon footprint over a model’s lifetime will then decrease over time. There are also advances regarding the most efficient way to train models that will be contributing to lowering the carbon footprint. One example is Neuromorphic computing, and it is an alternative that draws inspiration from how the brain functions. This technique is much less energy-intensive than current computer models and might be a solution to more sustainable AI.

Five quick questions with Mikael:

  1. Android or Apple?
    • I’m team Android. I do have to admit that Apple do make good stuff, but I still prefer Android.
  2. Movies or Books?
    • I’ve always loved books and used to read a lot. But I love movies too (I’ve probably seen more movies than what is healthy)! But usually, the book is way better than the movie. But I watch more movie than I read books today. I can’t choose!
  3. City or Countryside?
    • I prefer to be close to nature, so Countryside would be my choice. I especially love the sea and absolutely love to be in the archipelago.
  4. Messy desk or clean desk?
    • I’d like to think I’m quite neat and tidy, but in reality, I’m more on the messy side. I really do try to keep my desk tidy, but I’m not always succeeding though.
  5. Glass half full or glass half empty?
    • The glass is always half full! I’m an optimist by heart and I wouldn’t want it any other way. There are positives and possibilities in everything!

Tell us a fun fact about yourself!

  1. I am a hobby addict. But only for periods. When I find something that I truly enjoy, I go full on nerd. I read, watch videos, join forums, etc. to learn all about it. When I reach a level where I either understand I’m not going to be very good or that I must dedicate more time than I want to become really good, I quit and find something new. Currently, I’m deep into Disc Golf.
  2. I accidentally deleted a whole user database in one of my first employments. It was a good way to test our restore capabilities…
  3. The most “interesting” start to a company I had was after two weeks of training, I was sent to Boston, USA, to meet one of our prospective customers. I was there as an expert at our solution/API. My knowledge was drained after the first day, so I had to take all the questions during the day, and in the evening call Sweden to get the answers from the developers, code an example and bring back next day. Did that for 5 days, then I flew home on Saturday, flew to Bangalore, India, on Sunday and did the same for 5 days there with another prospect. When I came home, I was violently ill from food poisoning and jetlag.
  4. I really thought that the Windows phone had a bright future. I still stand by that! It actually was a good phone!

MainlyAI hires Mikael Nordenstjerna as Chief Executive Officer

Stockholm, Sweden, & Palo Alto, CA, March 6, 2023 — On the 6th of March 2023, MainlyAI, a Stockholm, Sweden, and Palo Alto, CA, based deeptech start-up, excelling in developing a shared economy for Artificial Intelligence, has announced that it is enhancing its journey by hiring Mikael Nordenstjerna as the company’s CEO.

Mikael brings vast experience in the technology field, with several senior leadership roles, including CEO and Country Manager. These experiences will allow Mikael to continue the introduction of MainlyAI’s innovative solutions to the ever-growing AI market, influence management, development, and upbringing of strategic opportunities, grow the network of users & partners, form corporate alliances, oversee further technology development, and raise seed capital for further expansion in addition to $1M of soft funding already received from the government innovation body in Sweden, Vinnova. Mikael reflects over his new role by saying: “I am thrilled to be part of the MainlyAI team. Latest developments of AI technologies are rapidly changing the world. My entire career has been spent in business development roles, and I look forward to continuing that tradition by improving adoption of these technologies by MainlyAI’s current and future customers.”

About MainlyAI

MainlyAI AB is a Stockholm, Sweden, and Palo Alto, CA, based deeptech research and technology company with the objective to allow businesses to work and cooperate on AI projects internally and externally. By sharing, selling, and acquiring machine-readable insights in a safe and privacy preserving manner, customers can improve accuracy of their AI models and optimise business KPIs. The approach of MainlyAI is centred around a platform as a service with an API providing a knowledge database of insights, and services simplifying the data/insight access and adoption of AI technologies for business.

Summary of 2022

In the year of 2022, more data than ever has been created, more privacy concerns than ever have been raised, and more AI-companies than ever have been founded. MainlyAI is one of many AI-companies on a mission to create insights from data in a trustworthy fashion, and we can see that even all actors’ efforts put together is currently not enough. Therefore, in the coming years, we look forward to increased AI efforts and to play a role in boosting collaboration in AI-applications across industries.

MainlyAI was founded in mid 2020. In 2022 we:

  • continued to contribute to the development of AI for precision health, specifically focused on early detection and prevention of autoimmune diseases, such as type 1 diabetes. Together with our partners, Diamyd Medical, Leading Healthcare Foundation, Lund University, National Diabetes Register and Sahlgrenska University Hospital, and with support from Vinnova, we have continued to develop and strengthen the innovation milieu ASSET.
  • continued to contribute to the use of AI in sustainable production through the projects ALISTAIR and EXPLAIN, with support from Vinnova and Formas, and in collaboration with KTH, Diamyd Medical, Scania, Uppsala University, Seco Tools, and AstraZeneca. 
  • provide AI expertise and solutions to the space company ISAAC that helps companies to transform existing space solutions to sustainable solution on Earth.
  • released our tool MIRANDA – a platform-as-a-service for design, execution, collaboration, and organisation of AI artifacts and projects.
  • continued to build a strong team in Stockholm, Linköping, Vilhelmina, and Palo Alto/Silicon Valley.

During 2023, we look forward to onboarding many new projects and partners to our AI collaboration platform MIRANDA, welcoming new employees, and investors to MainlyAI, and to contribute with AI solutions in our current domains of precision health, sustainability and space, and to explore new application areas. Employees, partners, and investors: Thank you for all the fun we had together in 2022. Let us make even more AI collaboration in 2023!

MainlyAI receives order from I.S.A.A.C.

MainlyAI has received an order from the space tech company International Space Asset Acceleration Company AB (I.S.A.A.C.) to deliver an advanced AI-driven database.

For more information, see I.S.A.A.C.’s press release

Monetarisation of AI

1. Status quo

Adoption of Artificial Intelligence (AI) technologies has been relatively slow so far. This has been changing rapidly in the past two-three years. However, the question of what the main hindrance for AI implementation remains. A couple of areas could be highlighted:

  1. Difficult to measure Return on Investment (ROI);
  2. Long time to market.

Let us analyse each point. AI can be used for making existing processes more efficient or developing new products. However, the below reasoning is valid for either scenario.

  1. Firstly, how could one assess economic returns from investing in AI? One way could be by monitoring company’s KPIs. If those are improved due to implementation of AI then one can measure Return on Investment (ROI) that has gone into implementing AI.

However, one only derives economic value from AI through deploying the models into production and not keeping the work on a purely academic level. This has been confirmed by findings in several reports, such as McKinsey[1] & Deloitte[2], that found that progressive AI practices are rewarded. Companies seeing the biggest increases in earnings from AI were not simply following common practices, such as establishing machine learning operations, MLOps, and IT process automation (AI for IT operations, or AIOps), but also spending more efficiently on AI and taking a greater advantage of new technologies. Secondly, the time to market. Entire departments might need to be set up and developed to identify usable data streams, create, train AI models, etc. On top of that, once a new technology is adopted, there is a dip in KPIs and performance. At times the length and the depth of such a dip impedes adaptation of new technologies.

2. What’s missing for improved KPI optimisations?

A lot has happened in the past years. However, there is, of course, substantial room for improvement in the economic returns from investments in AI. This can be achieved via:

  • Improvement of accuracy of AI models
  • Wider availability of models

2.1 How can improvement in KPI optimisation be achieved?

Introducing MIRANDA, which allows to exchange insights in a privacy preserving fashion to improve accuracy of models.

The number of AI models developed world-wide is, probably, approaching the infinity at this stage. This is no news. What is the crazy part in this? Many great initiatives get shelved or are kept at a purely academic level and don’t get a chance to reveal their full potential.

This is painful and frustrating for any company, especially for smaller ones with limited resources that need to adopt latest technologies as a core part of their business: medical companies looking to develop new and personalised treatments, fintechs looking to keep and expand their customer base through minimising customer churn… The list could go on, but you get the point.

To help them overcome these challenges, MainlyAI set out to build a platform that enables companies across industries to share insights across companies and even drawing upon insights from other industries. We can take the opportunity to name one concrete example here on how one can capture opportunities to learn from insights from other companies or adjacent industries. A drug company can apply insights from one type of autoimmune diseases, e.g. Diabetes type I, on the autoimmune disease it is researching, e.g. Rheumatoid arthritis.

We facilitate and democratise adoption of AI technologies, so any company can focus on their core business instead of building out support tools. The first indications for the future of such a platform have been very positive. MainlyAI has already been granted government funding for several high-profile projects.

Pardon, the obvious bias, but we think this is the next really, really big thing. It could be, in line with what Google did for search engines or Apple for app distribution. As crazy as it might sound, we are happy to explain our thesis, based on our own AIStore ™©, which allows storing and improving your existing models and purchasing trained or untrained models, off-the-shelf.

Should you like to join us on democratising AI in the society and help some very innovative companies in their continued growth along the way or, perhaps, you work for a company that is currently considering implementation of new technology – please reach out, we’d love to hear from you.

Regardless – if you are interested in learning more, go check out & follow us on LinkedIn.



Summary of 2021

Please watch or read about some of our highlights in 2021.

Looking back at 2021, the first full year of MainlyAI operation, we would like to express our gratitude to our employees, partners, and investors for their contribution to help us approach our vision – To turn data into knowledge and actions, using artificial intelligence.

In 2021, we strengthened the development team that now has competences ranging over UX, data science, AI modelling, database design, and cloud computing. Many thanks for your solid contribution to developing our MIRANDA platform – it is constantly evolving, and today we are ready to show the world its capacity and unique features.

Our already strong project portfolio has been extended with an innovation milleu funded by the Swedish innovation agency, Vinnova, that will benefit those risking to be affected by Diabetes type I and, in the longer run, other auto-immune diseases. We look forward to collaborating in a partner constellation consisting of Diamyd Medical, Leading Health Care, Lund University, National Diabetes Register, and Sahlgrenska University Hospital. The project, named AI for Sustainable Prevention of Autoimmunity in the Society (ASSET), will run over five years with a total budget of 60 MSEK.

We received support from KTH Innovation, and upon completion of the KTH pre-incubator program we partnered up with Busybee, a Stockholm based tech startup that offers occupancy detection systems and venue management to businesses. Finally, we would like to extend a special thanks to our growing network of partners: Astra Zeneca, Busybee, Diamyd Medical, Hitachi ABB, KTH, Leading Health Care, Lund University, National Diabetes Registry, Sahlgrenska University Hospital, Scania, Seco Tools, and Uppsala University.

MainlyAI and Busybee form a partnership

Just in time for Christmas, MainlyAI and Busybee have announced their intention to work together in a partnership to leverage on each other’s know-how and platforms in order to scale up their operations. To begin with, Busybee will stream anonymised data into MainlyAI’s MIRANDA platform, where insights and predictions will be created using artificial intelligence based on occupancy data. As the next step, the participating companies will be able to apply optimisation algorithms on the data streams. Being part of the MIRANDA platform opens up the possibility for privacy-preserving insight-sharing among enterprises to achieve higher precision in predictions and decision-making in shorter time.

MainlyAI in Innovation Milieu ASSET

MainlyAI has, together with Diamyd Medical, Leading Health Care, Lund University, National Diabetes Register, and Sahlgrenska University Hospital, been granted funding by the Swedish governmental funding agency, VINNOVA, in the Innovation milieus in precision health 2021 call. The project, named AI for Sustainable Prevention of Autoimmunity in the Society (ASSET) will run over five years with a total budget of 60 MSEK (40 MSEK from VINNOVA).

The aim of the innovation milieu ASSET is to contribute with personalised precision prevention of autoimmune diseases in the society. In particular, artificial intelligence (AI) techniques will be applied to learn from existing data to identify (i) individuals at risk of developing type 1 diabetes, and (ii) individuals with upcoming/newly diagnosed type 1 diabetes that would benefit the most from precision prevention or early intervention with therapeutic approaches. In doing so, data from the TEDDY (The Environmental Determinants of Diabetes in the Young) study, and other sources will be used.

The ASSET innovation milieu will be coordinated by Diamyd Medical. The application was allocated full funding amount by VINNOVA in strong competition. A complete list of the funded innovation milieus can be found at (in Swedish).

About MainlyAI
MainlyAI AB (559258-7538) is a research and technology-based company with the objective to allow for businesses to share data and insights in a safe and privacy preserving way and hence speeding up and democratizing the introduction of AI technologies. The approach of MainlyAI is centered around a platform as a service with an API providing a knowledge database of data and insights, and services simplifying the data/insight access and adoption of AI technologies for business.

AI Regulation, our take.

Last week, EU has published a proposal for an AI regulation, outlining risk classification for various AI applications, and proposed measures to be taken for each risk category. This came as a follow-up on EU guidelines for trustworthy AI suggesting stricter policies. Discussions in the media have been extensive since then, we at MainlyAI have been monitoring them with curiosity and excitement, and now it's time for our 5 cents.

AI has been proven to be an efficient technique in many industries. And as with many other technologies, it is important to regulate its use to guarantee the trustworthiness. When you download software to your computer or phone, you want to be sure that the software does not cause unwanted behaviors. This is even more important when we talk about safety-critical industries such as the automotive industry or medtech, because one must be able to rely on the software. Therefore, there are certifications for software at different levels. And, in a similar way, one should be able to know that AI algorithms do not add unwanted behaviors in the contexts in which they are placed. If a model has been trained on incorrect data, the algorithm will provide incorrect decision support and such situations can be avoided by regulating the use of AI at different levels.
One of the advantages of the proposal is that by regulating the introduction of AI in industries, one can, strange as it may sound, promote innovation. Without regulation, high-risk industries cannot rely on third-party developers. If you have a way of certifying an algorithm and guaranteeing that it does not cause unwanted behaviors, you can be safe when using it. Another advantage is that the algorithms around us will be reliable, and society will have greater acceptance of AI. 

The regulation may, however, lead to lower efficiency of the algorithms and somewhat slower market introduction for the AI ​​solutions in Europe. But no matter how much we love AI and efficiency in reaching the desired KPIs, we do think that trustworthiness in all its forms - safety, privacy, security, transparency, explainability, non-bias, and non-discrimination - is of the highest importance and must be treated as a hygiene factor. Let's embrace it.

MainlyAI in Production and Logistics project EXPLAIN

MainlyAI has, together with AstraZeneca, Hitachi-ABB, KTH, RISE IVF, Scania-CV, SECO Tools, and Uppsala University, been granted funding by the Swedish governmental funding agency, VINNOVA, in the Production 2030 call. The project, named Explainable and Learning production & logistics by Artificial Intelligence (EXPLAIN) will run over three years with a total budget of 13 MSEK.

The aim of the EXPLAIN project is to increase the profitability, sustainability, and competitiveness of the Swedish manufacturing industry by an innovative combination of virtual production technologies and AI algorithms. EXPLAIN will explore this combination as a unique way to provide increased access to new knowledge and skills in the production area, which is not possible with the current industrial practice of applying virtual production technologies alone.

The EXPLAIN project is coordinated by Uppsala University. The application was allocated full funding amount by VINNOVA in strong competition, with the success rate of less than 20%. A complete list of funded projects can be found at (in Swedish).

About MainlyAI
MainlyAI AB (559258-7538) is a research and technology-based company with the objective to allow for businesses to share data and insights in a safe and privacy preserving way and hence speeding up and democratizing the introduction of AI technologies. The approach of MainlyAI is centered around a platform as a service with an API providing a knowledge database of data and insights, and services simplifying the data/insight access and adoption of AI technologies for business.

For more information email