Machine Learning

Dr. Matthias Hirtschulz

Dr. Matthias Hirtschulz


Machine learning has a proven track record of solving complex problems on large datasets including unstructured data. Increasing computing power and the explosion of available information permits the use of learning algorithms to explore new business ideas, as well as improve existing processes.

The usage of machine learning has expanded greatly from its specialised beginnings in artificial-intelligence research and digital-native tech companies.

The financial service industry is no exception, and is rapidly adopting these techniques in fields as diverse as trading strategies, risk assessment, fraud prevention or better understanding customer needs.

Generally speaking, we see four key ways in which machine learning can be applied:

  • to automate processing,
  • discover previously unknown relationships,
  • make informed predictions and
  • augment human access to information.

In many cases, however, identifying suitable data sources, reducing the number of fragmented data silos and achieving a better interconnection between different data pools might be sufficient as a first step. An intelligent combination of internal and external data often allows for major improvements.

How to get started? d-fine next helps you make the most out of your data by bringing together cutting-edge data science and deep domain knowledge.

Overview of our services

Setup and consolidate data sources

  • Perform structural analysis for identifying and visualizing communication patterns or concentration risks
  • Combine structural data with performance data, e.g., a network of customer relationships with loan or stock performance, in order to improve decision making processes
  • Automating data-quality or entity-linking processes
  • Searching for information using natural language queries
  • Setup a self-service business intelligence analysis system for agile reporting


Statistical or machine learning models

  • Sentiment analysis of news for corporate ratings
  • Monitoring patterns of trading and communication activity for fraud detection
  • Uncovering patterns in the distribution of risk for improving collateral management
  • Analysis of payment information for up- and cross-selling opportunities