AI. Innovation. Everything in between.

Automated search and summarization of thousands of studies on protein behavior, receptor action, or metabolism.

Automatic reading, allocation and further processing of delivery notes, bills of lading and invoices.

According to Leufkens et al. (2012), typical colonoscopy screenings miss 22% - 28% of polyps and 20% - 24% of adenomas. Using machine learning, the live 2D image from colonoscopy can be reconstructed into a 3D model of the large intestine, which highlights segments that require additional coverage.

Predictive models for bed utilization, emergency room figures or staff-intensive services improve planning & shift management.

Early identification of delays, bottlenecks, or data inconsistencies using supervised and unsupervised learning processes.

Risk analysis and dynamic developments in the investment and capital market sector — with the help of predictive models based on daily updated parameters.

Computer vision to identify and classify packages, pallets, or damaged units.

Computer Vision recognizes products via camera, e.g. at self-service checkouts, for theft prevention or inventory automation.

Automated, verifiable and faster coding of ethics applications, study protocols, reports, or AE documentation.

Virtual sensors measure production parameters without additional hardware. Machine learning uses existing data, such as temperature or current values, to derive new, valuable variables.

The famous “Enhance!” from CSI and other TV series is no longer a joke. With the new generative adversarial network, the image resolution of one or a series of lower-resolution images can be improved.

Forecast models identify risks due to delays, bottlenecks or border stoppages and dynamically suggest alternatives.

More efficient processing of account opening documents or asset report data through natural language processing (NLP).

Conversational AI for drivers: NLP models recognize intentions and control on-board functions depending on the situation — even offline.

Reliable forecasts of electricity prices, trading volumes or customer demand on an hourly, daily or monthly basis.

Using camera and deep learning, seam defects, paint inclusions or component deviations can be detected and documented at an early stage.

sports telemetry. By tracking and analyzing player movement in the game, we can generate crucial contextual insights from data to show you the things you've seen but haven't seen. The findings help teams make smarter decisions, but also broadcasters and the media to bring their audience closer to the game with deeper analysis.

Monitor user-generated content to prevent content that violates your terms of use. Prevent behavior such as bullying and revenge porn that harms both users and the platform's authenticity.

Granular satellite images can be combined and used for geoforecasts

Cobots are flexible robots that learn by imitating the behavior of human workers. They provide a flexible method of automation.

Automating manual processes in manufacturing or logistics using advanced robotics