We are Artificial Intelligence Engineers

We make your AI learn on its own

We don’t teach machines and we don’t program software for them – we make them learn on their own. For this, we use modern Deep Learning through convolutional, adversarial, recurrent or simply transfer-learned Neural Networks. Our secret sauce lies in our experience of 20+ years in Machine Learning. Plus in our internal scientific research, for which we cooperate with ETH Zurich and Microsoft Research. The result is a vast amount of tools and pre-trained models that we can pick off our shelves. For use cases such as these:

• Healthcare
  Diagnostic image interpretation (classification, quantification, prediction)
• Manufacturing
  Quality assurance, Productivity (photo, video and scan-analysis)
• Retail
  Customer tracking, Sales prediction (video and sales data analysis)
• Insurance, Finance, Legal
  Document analysis and routing (natural language processing)

You train your own AI with your own Data

We provide and manage your AI platform, enhancing your AI success with our methods and tools.

• Project and Data Manager
  Usage Tracking, Auditing, Validation, Export/Import
• Image Preprocessors
  Resolution, Sizing, Chunking, Centering, Normalization, Gamma-correction
• Data Augmentation
  2D, 3D, GAN
• Keypoint Setting
  Image Segmentation
• Deep Learning and Storage Platform
  Microsoft Azure
• Support Services
  Coaching, Training, Model Optimization

Automated Document Processing trough NLP

Classifying, Routing and Acting on Content in Unstructured Documents

• AI-driven, semantic text content analysis
• Automated extraction of classification information and codes
• Apply storage location, recipients, target processes and other attributes automatically
• Integrate with Microsoft Office 365 and Microsoft Teams

Our Investment for better and less expensive Healthcare

Detecting and measuring clinical conditions in Musculoskeletal MRI

• Automatic interpretation of musculoskeletal MRI images.
• Use cases: Real-time second opinion, peer review, retrospective quality review, case comparison, etc.
• Large catalog of detectable conditions (frequent and rare ones) currently being trained in cooperation with leading Swiss hospitals.
• Prototype is in clinical trial (retrospective) since September 2018, finalization expected 2019