We are only in the very first phase of AI disruption in the financial sector and, in general, in technology. At the moment, progress is limited to environments where large quantities of historical data are available and often relates to ways to automatize and improve human sensory capabilities (vision, natural language processing and outliers’ detection). The next big jump will be to move beyond the improvement of human capabilities to new ways to leverage artificial intelligence. This will be done by accepting the fact that explicability of AI predictions is mostly detrimental to their performances, and developing and deploying new techniques will allow AI to learn even from little data.
One such company named Axyon AI brings deep learning-powered solutions to finance. The company’s technological advantage is represented by a proprietary deep learning platform specifically built for financial problems. The platform allows Axyon to be extremely quick in developing highly-accurate predictive models. Axyon developed successful POCs in several fields, from credit risk to wealth management, from churn-rate prediction to fraud detection. It currently sells two products: SynFinance, for the loans market, and StocksAnalyst, for AI-powered asset allocation.
Fascinating Journey of Axyon AI
Axyon stems from a three-year long R&D effort in applying AI to finance, going beyond its most common applications. Axyon was spun-off from its parent company in 2016 and received two rounds of investment from ING Bank. They aim to bring AI, and especially on the platform of deep learning, into very specialized and high-impact niches of the financial sector, beyond the common use cases that are targeted by most of the companies. In the long run, they also aim at enlarging the scope of their efforts, using the acquired knowledge to bring AI in human welfare-related fields that are currently overlooked by AI advances.
Applauded Moments of Axyon AI
Axyon received various press coverage in the last few months, being mentioned by ING’s CEO in a global conference in September as an example of bank/start-up partnership. Last year, Axyon was selected in several competitions, reaching finals in EIT Digital Challenge and BBVA Open Talent, to name a few.They pride of having always surpassed benchmarks provided by the clients with their deep learning models.
Backbone of Axyon AI
Daniele Grassi, CEO, Axyon AI is a 10-year long IT entrepreneur, TEDx speaker about AI, passionate about artificial intelligence, human learning, and science in general. He laid the R&D groundwork for Axyon technology, and now in charge,he is handling the responsibility of directing the technological effort of the organization. Daniele has always tried to see before others the potential for applying leading-edge technology to real applications. He also had the luck of staying in contact with great professors at the University of Modena (a special mention goes to Simone Calderara, Ph.D.), who always kept the team Axyon in touch with the most recent advances in the AI field. Daniele loves to experiment new techniques as soon as they are out, often in sprinted experiments that are the love (and sometimes the disruption) of R&D team.
Axyon is one of the leading examples of AI companies in Italy. It has strong connections with the University of Modena, home to (arguably) the strongest research group in the country for AI and deep learning.
Challenges faced by Axyon AI
Axyon faced several challenges along the journey. The very first one was the catch-22 scenario of having to show a good track record in applying AI to finance to be able to access the financial sector. The company overcame this thanks to targeted partnerships and acceleration programs. Understanding human diffidence versus deep learning predictive models (commonly referred as black boxes) became their first enemy. Axyon AI challenged this with a targeted R&D effort in techniques to add a layer of approximate explanation to their models’ predictions.
Future Executions of Axyon AI
Axyon is growing at a steady pace, and Daniele sees it as being a market leader for AI applied to high-margin niches in investment banking and asset management in three years. By that time, hype for the first push of AI will be already decreased, and only companies that have aimed to grow with solid technology and solutions will be there to lead the following AI charge. Axyon will be one of them.