We are proud that we can offer advanced AI applications in the financial service industry and have been recognised for our solutions.
A recent report “The State of AI 2019: Divergence” by MMC Venture found that only “in approximately 60% of the cases – 1,580 companies – there was evidence of AI material to a company’s value proposition”. That means that around 40% of European startups in that study use the term AI most likely only for marketing purposes. AAAccell, however, rightfully brands itself as being an AI company. Last year the AAAccell was selected from the Google for Entrepreneurs Exchange program, organised by Impact Hub Zurich, as one of ten Artificial Intelligence and Machine Learning startups from all over the world. We are proud that we can offer advanced AI applications in the financial service industry.
What is Artificial Intelligence (AI)?
We understand AI as the summary of a science, aiming at simulating intelligent behaviour in computers.
In this context, AAAccell does not use AI to describe replicating independent human intelligence. Instead, we apply the term to define a set of systematic methodologies enabling algorithms and computers to exhibit a subset of human behaviour. In particular, we focus on the capability of learning, pattern recognition, and decision making, as well as problem-solving.
From where does modern AI in this context originate?
A common consensus is that the current understanding of Artificial Intelligence and its computational implementation can be traced to 1955 and “A proposal for the Dartmouth summer research project on artificial intelligence” by John McCarthy – an American computer scientist. The goal of the motion was the following:
“The study is to proceed on the basis of the conjecture that every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it. An attempt will be made to find how to make machines use language, form abstractions and concepts, solve kinds of problems now reserved for humans, and improve themselves.”
The research proposal and its findings had lasting implications, and John McCarthy went on to develop Lisp, a programming language still in use today. Nowadays, various programming languages including R and Python allow for solving different types of problems. One common method falls under the name Machine Learning (ML), which consequently is a branch of Artificial Intelligence.
What is Machine Learning?
AAAccell understands Machine Learning as an algorithm or a set of algorithms learning from examples and experiences (i.e., data sets), rather than only confining calculations. For example, instead of telling a machine how to differentiate between a bar of chocolate and a piece of cheese, an algorithm is fed with large amounts of data and trained to allow the computer to learn on its own how to distinguish between the different types of objects. This image recognition task is an example of a “classification algorithm” since it classifies different kinds of input data.
How does AAAccell use Machine Learning?
A compelling example of an algorithm that can be used for ML classification tasks is the “Random Forest Algorithm” – a methodology widely tested in academia for its promising applications in finance. AAAccell uses different versions of Random Forest Algorithms because it is possible to employ the method for both classification and regression problems. One such implementation that we developed is a regtech solution to calculate a risk profile for illiquid assets including real-estate funds – a problem that regular financial models cannot achieve. We employ our proprietary Random Forest to build a portfolio that replicates the movements of illiquid assets along a time series. We base our calculations on worldwide data on asset classes, market instruments as well as a wide variety of economic variables. For this solution, we were selected as a RegTech100 company in 2019.