The Researcher Workspace
Creating a well-functioning tool for processing scientific and technical textual material using AI is no easy task, but over the past 7 years, Iris.ai have achieved just that with their main product, The Reasearcher Workspace, which offers researchers with a number of smart tools to simplify the way they process, find and use scientific data material. The core of their machine learning engine is trained on a significant amount of scientific articles, and when it comes to Natural Language Processing, Iris.ai is deemed to be one of the best. The service can be tailored to meet specific research needs, even when it comes to narrower subject areas making use of open access literature, as well as allowing you to add an unlimited amount of external sources. The consequences of simplifying these perhaps tedious processes in research can be extensive, and we applaud Iris.ai on their groundbreaking work.
We were lucky enough to have a quick word with the founder and CEO of Iris.AI, Anita Schøll Abildgaard, at our CuttingEdgeAI event about Large Language Models earlier this month, where she did a keynote presentation about building computationally efficient AI that gets the facts right. Iris.ai is one of the very first members of NORA.startup and Anita herself is a member of the board.
How will this grant impact your work going forward? What will be your main ambition in 2023?
The simple answer is that we can put active fundraising on the shelf for a while and focus on the goal of the EIC project: completing 'the Researcher Workspace'. Some new AI models need to be integrated, we need to improve the user interface and carry out user testing - as well as work closely with current and new customers to simply create the best research tool in the world, built on smart language models where facts are the most important !
Can you tell us a little bit about the EIC application process?
We have applied for the EIC many times before. It is a long and hard process, and when we reached the final round and interview last year and were rejected, we felt quite disillusioned and exhausted. But we had received the Seal of Excellence, the opportunity to apply directly to the 2nd step with a rebuttal - so when the final opportunity for that was approaching, we gave it one last try, and we are so glad that we did!
What would you say to other Norwegian startups that are considering applying for funding through the EIC Accelerator? Do you have any hot tips or inspirational words?
Haha, I don't really know - it's often strenuous and frustrating. Be absolutely sure that you fit into the category - i.e. innovative project that cannot easily get venture capital - and don't expect to succeed on the first try. If you use consultants, which many unfortunately find themselves having to do, then make sure that you have full ownership of the entire application. Accept all the help you can get, for example from Innovation Norway, which can both help with information during application writing, and provides free coaching and a general test for all Norwegian companies that come for an interview. And, don't put all your eggs in one basket - the chance of getting the EIC Accelerator is so small that it cannot be the only or most important strategy to get funding!
Do Iris.Ai have any exciting news or future plans that you are willing to share? We would love to hear it!
Well, I can say briefly mention that in addition to our two-year project as part of the EIC Accelerator, we are also working towards achieving a big goal, which is where the European Investment Bank's co-investment of up to 120 million comes in: what we call an "AI researcher", or more specifically a "scientific question-and-answer machine", which will be able to keep track of facts and scientific knowledge, and answer all questions related to research, whether the answer is found directly or inferences must be made based on multiple sources. We see that current large language models such as ChatGPT, or the failed Galactica from Meta, give convincing answers that are completely wrong and made up - and this is precisely where we believe that smaller, laser-focused companies that conduct their own research, like us, can succeed. We are working towards becoming market leaders in that artificial intelligence keeps track of the facts, in short.
Learn more about Iris and the Research Workspace on their website, www.iris.ai
Interested in learning more about NORA.startup and in becoming a member? Contact birte.hansen@nora.ai!
This articles was originally published on the 13th of March, 2023.