Companies in business sectors ranging from Technology to Life Sciences, Automotive and Finance, are competing to build top-notch AI teams to transform their business, and to recruit the best global AI talent, but are these investments productive? With millions of people worldwide working with AI algorithms in 2020, and over 90% of mid-size and larger companies now having a specialized AI or Data Science team, researchers and engineers in this field are literally drowning in the pace of innovation. A lot of the leading innovation by large research teams at Big Tech companies is published in open academic research platforms, in blogs and accompanied by open source software, but the amount of information is starting to exceed human capacity. 

 

Leading researchers in the field of AI, like Yoshua Bengio and Yann LeCun, openly admit they find it almost impossible to keep up with the latest developments. Per day, an AI expert needs to scan and digest several hundred new research publications on open publication platforms like arXiv to stay up to date. And that does not count blogs and news. That is a lot of very expensive hours.

 

Amsterdam based startup Zeta Alpha is now launching a new deep learning based search platform, called AI Research Navigator, to help AI and Data Science experts organize this flood of information. Zeta Alpha founder and Dutch AI veteran Jakub Zavrel, previously responsible for the successful recruitment AI scale-up Textkernel, started developing the tool to apply the recent advances in language understanding by deep learning networks to the information overload in his own field of expertise: “I never understood why AI people do not take their own medicine for organising their research”. He believes that AI has the potential to help people make better decisions in their areas of expertise and therefore do better work. “General academic search engines like Google Scholar, Microsoft Academic or Elsevier Scopus do not have enough AI specific focus, and popular open publication platforms in AI, like arXiv.org, do not have modern search and recommendation technology”, he adds. Zeta Alpha is currently working with top AI researchers at the University of Amsterdam, and with a number of other Amsterdam based AI startups as launching customers of the tool. With an experienced international team, Zeta Alpha is looking to capture a large global user base with this specialized information platform.

​

While everyone can witness the acceleration of technical progress in deep learning models, e.g. in the ability of the recently introduced OpenAI GPT-3 model to generate human level writing, the experts are divided on how the AI field will evolve. Will AI split into many specialized sub disciplines, or will it maintain the current cross-disciplinary nature and the fast pace of development? “It depends on the way we can use AI technology itself to support the researchers in the field” says Marzieh Fadaee, who recently completed her PhD research in AI at the University of Amsterdam is now a research engineer at Zeta Alpha.

 

The last thing companies investing in AI want is that their highly paid expert teams miss the boat by betting on the wrong AI trends or technologies, because they don’t have time to keep the flow of new research under control. AI is here to help, and to mitigate that risk. Next year, Zeta Alpha also plans to help teams navigate their own private information sources. Hiring the right people and finding out what the competition is up to are additional planned extensions of the product.

 

About Zeta Alpha
Zeta Alpha is building a new AI Enterprise Search and Insights platform with the goal to help people make better decisions and organize their work. Zeta Alpha was founded in 2019, has eight employees, offices at the Science Park in Amsterdam, and close ties to the AI research community at the University of Amsterdam. It has so far received €1M in private funding.

​

​

​