Speaker: Michael Berthold (KNIME)
Data science requires not only a broad spectrum of skills but a solid understanding of the underlying methods and the ability to blend tools and technologies. Visual programming, if done right, provides the appropriate level of abstraction, access to the relevant nuts and bolts, as well as transparency and governance capabilities. The presentation also highlights the top five items a visual programming environment provides for the data scientist on the one hand, and the organization on the other.
Speakers: Hans Samson, Marlene Hakvoort (ANWB), Moderator: Mihaly Medzihradszky (KNIME)
How ANWB built a predictive marketing solution using KNIME Analytics Platform - featuring the KNIME Model Factory and Integrated Deployment - that is easily manageable, scalable, standardized, and fast. The result? It now takes four hours to score 51 models with on average three million customers per model. This is 20 times faster than before - and the models can be scored every day, automatically.
Speaker: Dennis Ganzaroli (Swisscom), Moderator: Jeanette Prinz (KNIME)
The goal was to predict how COVID-19 cases are evolving and answer the question - when will the pandemic end? Dennis Ganzaroli does this, blending Jupyter, Tableau, and KNIME as a powerful visual programming tool. He applies Loglet-analysis as it takes into account the evolution of multiple waves. The result is an interactive dashboard enabling users to select a country or continent and investigate the expected cases rates.
Speaker: Alzbeta Tuerkova (Uppsala University), Moderator: Jeanette Prinz (KNIME)
Biomedical information mining is increasingly recognized as a promising technique to identify novel drug candidates. Here, Uppsala University developed a KNIME workflow to identify existing drugs which could be effective for COVID-19 treatment. While working with the small molecule data, several challenges were faced - such as an abstract chemical representation and the heterogeneity of the data in the public domain. The application programming interface and the integration of cheminformatics libraries and external software in KNIME Analytics Platform enabled them to tackle the problems in a clear, efficient, and reproducible way. It’s worth pointing out that some of the identified hits have been found to currently undergo clinical trials.
Speaker: Madhu Raman (Amazon Web Services), Moderator: Paul Treichler (KNIME)
How the collaboration between KNIME and AWS is an exciting step in the world of intelligent automation and addresses the following pain points for industry customers on AWS; improving time to value, enabling workflow integration and low-code development, overcoming lack of machine learning and data platform engineering services, and getting the workflows ready for deployment, securely.
Speakers: Lexi Harrison, Olivier de Bie (Forest Grove), Moderator: Paul Treichler (KNIME)
Here the Forest Grove team easily blends tools to modernize a process. An interactive bidding system was developed with KNIME as a backend to three further tools: Microsoft PowerApps, Microsoft Flow, Swagger UI. KNIME Server Medium is used to deploy the app: A REST API was created and multiple GET and POST requests receive and send data. It was an excellent project to see how well KNIME integrates with PowerApps in particular, but with multiple other tools in general.
Speakers: Yentran Trang (BOSCH IT), Ioannis Magnisalis (European Commission), Moderator: Phil Winters (KNIME)
BOSCH has 125 manufacturing sites and 11 IT divisions, all pulled together by BOSCH IT. The European Commission relies on central IT services to handle their needs. In this panel discussions Yentran (BOSCH IT) and Ioannis (European Commission) discuss orchestration, reusability, security, and compliance in the enterprise and the importance of having a tool that provides the respective capabilities. Criteria for success? Find a tool that lets you concentrate on your solution rather than the tool.
Download slides from the breakout sessions: