Working with data inevitably means organizing it. Especially in advanced analyses with artificial intelligence (AI), the combining, preparation and cleaning of the data determine a large part of a analytical project. In order to achieve the defined analysis goal, different data formats such as streaming data, relational data and Excel files have to be joined. The merging and data preparation is ad-hoc and must withstand ever-increasing data volumes.
While data integration and preparation is also possible in traditional systems, these are mostly specialized for expert users. Modern data management platforms for AI projects, however, are a place for collaboration and bring diverse users close to data. Visual representations of processes, a flow supportive user experience, and a low-code mentality keep barriers to entry low and clear the way for self-service analytics.
The data management solution SAP Data Intelligence combines usability for beginners, functionality for experts and scalability of data processing in one system. The successor to SAP Data Hub with the integration of SAP Leonardo Machine Learning has been available to end customers since 2019 and is being continuously developed.
Unite different types of data with ETL or alternative ELT processes. Operational connections to SAP systems as well as non-SAP systems are available. Data from an Excel spreadsheet can be used in the same way as from an IoT stream.
The data is centrally bundled in the data management solution. An integrated data lake is available for heterogeneous file formats. Comprehensive metadata management and transferable rules to ensure data quality help with the necessary organization.
In the Modeler component, you can create any desired complex workflow for your data processes using drag-and-drop. The visual representation ensures transparency and a recognition value of the operatores and steps.
The data science process is optimally supported by an interaction of various components. With the ML Scenario Manager, the exploratory data analysis, the workflow and the models including the associated target metrics are provided in a versioned manner.
A built-in JupyterLab instance allows scripts in the programming languages R and Python to be integrated. SAP is thus opening up to the rapid change in open source technologies and providing advanced users with up-to-date algorithms.
SAP Data Intelligence emphasizes reusability. Automatic rules to ensure data quality and workflow parts, as well as user-defined operators, are easily transferable. Development time benefits from the very beginning due to numerous templates for diverse application scenarios.
With a strong backend of well-proven SAP technology in combination with open source systems such as Kubernetes and Docker, the technical realization of data processing is no longer a concern. Work is performed on a data extract and the overall processing is started at the push of a button.
The components of Data Intelligence are available in the modern and modular web interface in tile view.
The various functions are arranged in a user-friendly context in the respective sub-area.
Data Intelligence Launchpad
Modeler
Meta Data Explorer
Metrics Explorer
Data Preparation
The following third-party systems and external connections can be connected without any problems:
Thanks to the easily accessible web interface, low-code functionalities for the most important data preparation steps and the visual representation of the workflow design, business users quickly find their way around and implement their own analyses.
All phases of a data science project can be bundled in SAP Data Intelligence. Data, models and deployment all find their place here and are presented visually.
SAP Data Intelligence Cloud is available as a subscription or pay-as-you-go model and is delivered on-premise in your own data center, private cloud or via all major cloud service providers.
Due to the central access to different data sources, data ratings and commenting functions, Data Intelligence can be used as a creative space for individual data projects with AI content. Due to the low entry threshold, business users can be inspired to implement their own ideas. The Data Science Journey is flexibly guided by the structure of the ML Scenario Manager.
Your Machine Learning projects will benefit most from Data Intelligence. The bundling of project information in the ML Scenario Manager, the realization of complex workflows in the Modeler and the execution including extracted target metrics are only a few advantages worth mentioning. The versioning and transparency also convinces expert users.
In this article, the general advantages of license-based systems and an open source architecture are compared using the example of SAP Data Intelligence.
What can SAP Data Intelligence do? What can it not? Find out in our new whitepaper!
Join us for a brief look at the key features of SAP DI and its development history, as well as the potential future of the SAP Business Data Platform.
12/03/2024 - 03:00 to 04:00 p.m. (CET)
