Process

We divide the ML development process into three buckets.

1. Define the technical hypothesis, e.g. clinical problem, and translate it into the domain of AI.
2. Analyze the data and train the algorithms e.g. to classify the data/images
3. Automate complex diagnoses by incorporating previous results and additional information.

We focus on the first and third areas and are supported by competent partners for the middle area.

Using our experience and that of our partners, we support to create intelligent agents and decision support systems. Our software agents, whether automated players in computer games, for medical diagnosis or others, receive tasks, plan the path to a possible solution and execute each step until they have a suitable solution. They learn from their environment and from user feedback. Decision support systems assist experts in their decision-making process for a given task by proposing appropriate hypotheses, validating them on an ongoing basis and calculating the probability of events occurring.

Projects

A brief selection of our successful projects

BOOST: AdaptaBle AutOmated Intelligence Gathering PrOceSses for Decision SupporT” – 2023/2024 FFG funded – in cooperation with the AIT Austrian Institute of Technology, Syncpoint GmbH for the Bundesministerium für Landesverteidigung
The aim of the project is to support analysts by automating individual steps in the intelligence process. The first step is to structure the collected data, the second step is to extract relevant information, and the final step is to derive suggestions and validate hypotheses through automated reasoning. More information about the used platform is available at syncpoint.io/isr.

“PIONEER: InteroPerability and DIgitization Of INtelligencE GathEring Processes” – 2021/2022 FFG funded – in cooperation with the AIT Austrian Institute of Technology, Syncpoint GmbH for the Bundesministerium für Landesverteidigung
As a predecessor of the project BOOST the focus of this project was to support the intelligence analysts by automatically and manually structuring of the collected data and the automatically construction of the knowledge base using AI methods, supported by a dedicated user interface and digital tools. The knowledge base was used as a source for presenting the information in a network diagram as well as in time-based and geo-referenced images.

Recognition of hand-drawn military symbols for inclusion in the digital situation map.

Several projects for the game developer Greentube in the area of network layers, graphics and NPC agents.

Production control system for Oracle database to derive daily closing prices of assets for the company AIM.

ASP Solver in D and it’s application using custom api e.g. SQL on various use cases e.g. route planing, simple medical diagnoses.