Our research is focused on getting machine learning into production systems. and to do so in a responsible and sustainable way. We operate at the forefront of Machine Learning Engineering (MLE). Based on our expertise, we work together with you on scaling, testing, data modelling and other major MLE specific aspects such as operations and maintenance. When we build these applications, we measure success based on business relevant metrics. Members of our team have been responsible for launching some of the largest Natural Language Processing applications to date in the U.S banking sector, and their work has been covered by NBC Universal, SRF, The Economist and the Handelszeitung.
Launching Smart Applications
We cooperate with industry, academic research groups and non-profit organisations to build applications that can offer novel functionalities or expand beyond existing capabilities based on machine learning.
The following domains represent our current focus and are complemented by applications we developed in other ML areas such as computer vision and inference.
Smart Text Applications
Smart text applications are aimed at making the interaction with information in textual form more efficient and rewarding. The applications range from the identification of key values in important documents, to automated curation, transformation and storage of documents in all forms. The systems we develop with our cooperation partners automate partial aspects and reduce processing time by up to 80%. They follow the principles of intelligence amplification (IA) and empower humans, instead of aiming for the pipe dream of an AI lawyer or AI doctor. We currently work with some of Switzerland’s leading start-ups in legal-tech and HR-tech on building smart text applications.
Search & Recommendation Systems
Search and recommendation systems are key areas where advances in Machine Learning have unlocked enormous potential. Search not only has the potential to increase the efficiency of information-rich workflows, but also to make users happier. Improving search in an e-commerce context can lead to better conversion rates and customer satisfaction. We support our cooperation partners in the realisation of search solutions and recommendation systems that unlock the potential of internal information and enable them to launch data market focused solutions.
Accelerating Research for Social Good
We apply the power of ML based automation and information extraction to accelerate and expand research in the social domain. Our work ranges from scaling the collection of required data on public and private sources, to the automated transformation, extraction and interpretation of information. We develop innovative tools for clinical intervention in psychology, for example to support burnout detection, and investigate how smart technologies can be applied to support diversity in job ads. Such projects allow us to support social institutions, non-profit organisations and public administration to benefit from innovation and to get the most out of the digital age.
We target different types of collaborations with partners from industry, public administration and academia to make innovation happen and do our part of pulling Switzerland into the future.
Different forms of collaboration are possible:
Third-party funded research cooperation based on public research programs
Contract research where we consult, advise and run workshops
Support and collaboration with start-ups
Student projects / bachelor theses relating to MLE.
We are happy to help you identify the best suited way to fund research cooperation.