Background: Artificial Intelligence for Retail

There is no “official” definition for AI. Artificial intelligence supports people when they are working with large volumes of data that feature complex relationships. This data needs to be analyzed quickly before decisions can be made on the basis of that analysis.

An essential feature of any AI is that it is able to learn. Initially “the machine” learns with training data before generalizing its knowledge and applying it to unknown data (keyword: machine learning). Datasets change very quickly these days and “the machine” learns from every new dataset that it processes (keyword: incremental or adaptive learning).

Good decisions: Our intelligent algorithms always provide you with the best decisions, giving you valuable competitive advantages in a highly dynamic market. Depending on the area of application and strategic objective, we combine suitable algorithms to best optimize the target figure.

Real time: Our personalization and pricing solutions constantly analyze all relevant contextual conditions including customer behavior, inventory and competitor prices, adapting their behavior to changes in real time.

We often compare the way AI works in our personalization and pricing solutions to the way a chess computer works: The computer recognizes the “moves” of the customer, anticipates the chain of all possible next “moves” and develops its own actions (e.g. product recommendations) so that it wins “the game” (e.g. the target parameter of customer value – not only short-term sales maximization). Granted that greatly simplifies the situation but it also paints a picture of AI that even the layperson can understand. We would like to emphasize in particular the long-term optimization of the target parameter of Customer Lifetime Value across the entire customer journey (keyword: network chain optimization).

From a historical perspective, the development of our AI technology started with data mining. Along with the availability of large volumes of data came the desire to analyze ever larger datasets (keyword: big data), recognize patterns and use this knowledge profitably. Data mining continued to develop in the direction of machine learning, which far exceeds simple pattern recognition in datasets.

Following the learning phase, “the machine” generalizes its knowledge and is then able to apply it to unknown data. In the beginning “the machine” learned only from static datasets but there was soon a need to learn adaptively (or incrementally). When it comes to adaptive machine learning, “the machine” learns from every new dataset that it processes. There are considerable advantages to this when it comes to both speed (keyword: real time) and quality.

It is no longer possible to manually manage a customized approach to each individual customer via every channel with the “perfect” offer that corresponds to the customer’s current shopping behavior, is in stock or is available in the customer’s desired store. This is where artificial intelligence (AI) in the form of self-learning algorithms comes into play.

Our personalization solution learns personalization rules (keyword: adaptive learning) automatically and in real time – e.g. on the basis of clicks, shopping carts, purchases, external and internal search queries, clicked categories and banners or even selected events. The AI-based software collects all of this data independently and continuously. In addition, the algorithms measure the acceptance of the generated content (keyword: reinforcement learning). If a generated recommendation is accepted by the customer, the system rewards itself. If the recommendation is not accepted, there is no reward. This is how the recommendation engine learns gradually and can better cater to each individual user in a more targeted way with each interaction.

Translated to retail, this means that if you use an algorithm that works according to the principle of reinforcement learning, you will be able to maximize the benefit of your “whole system” over the long term – in other words, the crucial variables such as sales, turnover and profit.

If you would like to learn more about personalization, we recommend our page entitled What is personalization.

Jens Scholz, founder and current CEO of prudsys AG, said it best in an interview written up in our blog: “prudsys AG has been involved with artificial intelligence and the automation of decisions since its founding. […] We found the challenge of measuring data in real time, learning directly from that and being able to play back the results in real time to be particularly exhilarating.”

We have been involved with the topics of machine learning and reinforcement learning, and later with AI, since as far back as 1998. At the time, a group of math and computer science students at the Chemnitz University of Technology decided to apply the theoretical knowledge they had learned in data mining to practice (among them our founders Jens Scholz and Dr. Michael Thess). Over time they figured out the appropriate field of application for data mining – the retail industry. They used new ideas to find opportunities to automate and optimize processes, allowing them to follow today’s vision.