Contract research AI

Top interdisciplinary development team for digital transformation

Our research and development services include the integration of AI, IoT and advanced (data) analytics to give you a sustainable competitive advantage through digital transformation .

In cooperation with our medium-sized international partner company, which employs some of the brightest minds, including graduates from Stanford and Princeton, we solve the most demanding mathematical challenges at the interface of physics, mathematics, mechanical engineering, IT, electronics, optics, and artificial intelligence (AI). What makes us special are the interdisciplinary development teams, comprised of IT specialists, engineers, mathematicians, and physicists, who deliver the most challenging custom development work . Compensation is purely success-based; our clients only pay for development services if the predefined and agreed-upon development goals are achieved . Our partner has already completed over 100 projects in the following fields for leading companies (including a German automotive manufacturer): digital transformation (image recognition, pattern recognition, machine learning, artificial intelligence), medical devices, AI in medicine, robotics, consumer electronics, and the automotive industry. They also have extensive experience in hardware (IoT, cloud, sensors, analog) .

In addition to our partner company, we have now built our own international multidisciplinary team.

Example projects for contract development in the field of Operations Research / Digital Transformation

Logistics optimization : Addressing logistics challenges such as vehicle routes, container stacking, goods packaging, and transport planning is extremely complex. Despite their "NP-complete" nature, which indicates their immense complexity, our advanced algorithms and heuristic techniques enable solutions that come remarkably close to the optimum.

Parallel scheduling : Running algorithms on large supercomputers requires dividing tasks into smaller components, each executed on its own individual node. Our expertise in scheduling algorithms answers the complex question of which tasks should be executed when and where.

Optimizing public transport : Determining the optimal timetable for a given route or deciding on new routes requires algorithms that best serve the public interest, which is in itself a challenging task.

Pricing : Efficient product pricing requires forecasting multiple future pricing decisions, taking into account competitors' actions and potential economic changes. Techniques such as dynamic programming help in selecting optimal decisions across various future scenarios.

Optimizing oil and gas production : Our platform integrates data- and physics-based simulations for large-scale oil and gas production projects to run multiple scenarios and generate future forecasts.

Reference projects in Artificial Intelligence (AI)

Recommendation systems : Our developed recommendation systems maximize customers' limited attention spans by providing precise and rapid suggestions for relevant products. Furthermore, they help customers become more familiar with new products based on subtle need recognition – a hallmark of high-quality recommendation systems.

Pricing : We create pricing models for companies that utilize a balanced combination of techniques from economics, finance, management theory, algorithmic computing, AI, and data science. Our pricing models not only propose prices but also explain their proposals in a way that is easy for people to understand.

How does AI support pricing?

Pricing your products is one of the most challenging aspects of running a business. How can you price your product optimally? How do you account for inventory costs? How do you plan for market changes? How do you react to competitor price changes? We develop pricing models for businesses by using a balanced mix of techniques from economics, finance, management theory, high-performance computing, AI, and data science.

Human-Guided Machine Vision : Our HGMV system combines fast image processing and machine vision algorithms with an advanced user interface for instant human feedback. It delivers accurate and context-aware results in real time across many fields, including meteorology, medicine, film restoration, special effects, archaeology, and more.

Intruder detection in videos : We have developed algorithms for the real-time detection of intruders in video streams. These algorithms can be used for security purposes and to ensure safety in business premises, on railway tracks, public roads, etc.

Social graph mining : Our social graph mining algorithms help to discover concepts such as subject-matter expertise, collaboration graphs, connection recommendations, and behavioral predictions for viral marketing from connection and conversation graphs between people.

Example projects in the field of signal, image and video processing

Oil and gas production : Thousands of sensors in an oil field continuously generate data. This data can contain missing values, incorrect values, noise, and may be sampled at different times and rates. It is important to generate high-quality data before it can be used to better understand the physics of the underlying reservoir and the production network.

Optimizing business processes : Econometrics is the application of signal processing and statistical methods to economic data. Such techniques are important both theoretically and practically. In practice, signal processing and data science techniques can be used to create quantitative market models. This helps answer questions about pricing and inventory strategies.

Noise reduction for hearing aids : Hearing aids help older people and those with hearing loss to lead as normal a life as possible. To achieve the best results, the sound must be presented to the wearer with maximum clarity and minimal noise or interference. Furthermore, this must happen in real time.

Video extensometry : To understand the mechanical behavior of objects, it is important to measure stresses and strains. One way to measure strains is by using video processing techniques known as image flow analysis. Accuracies can be calculated across the entire deformed object.

Video Codecs : Our partner company has worked on many video encoders and decoders (codecs), including standards-based codecs such as MPEG and H.264, as well as custom video codecs for specialized applications. In some cases, we have leveraged the computing power of multi-core processors to increase codec performance while saving energy.