SAP offers a machine learning framework through its Leonardo platform. The tools is clearly designed for large enterprise customers, and is the best fit with existing SAP customers. The solution is designed to address complex AI and machine learning challenges that connect to specific business processes and tools, including SAP Cash Application, SAP Service Ticket Intelligence, SAP Customer Retention and SAP Predictive Analytics.
The framework includes deep learning and AI neural nets and advanced predictive modeling and capabilities. The machine learning tools also support SAP Conversational AI and SAP Leonardo Machine Learning Foundation. Leonardo Machine Learning Foundation includes extensive APIs for image and video, audio and speech, natural language and tabular and time data. The platform works with TensorFlow and includes ready to use services.
SAP users are likely to find Leonardo Machine Learning a good fit for their enterprise. That’s because it seamlessly integrates with data from other SAP systems and data repositories that, together, deliver practical intelligent applications with a real-world business focus.
However, Leonardo Machine Learning also supports a Bring Your Own Model approach through TensorFlow, Scikit and R. This delivers a high level of flexibility and customization. Forrester has recognized SAP Predictive Analytics as an automation-focused PAML solution, which is a rapidly growing market segment.
Because Leonardo Machine Learning runs in the SAP public cloud, it can tap GPUs to address compute intensive deep learning and machine learning tasks. SAP also includes built in business services and templates to facilitate data collection, model development, deployment and refinement.
Overview and Features
Data scientists, statisticians, data miners, business analysts, citizen data scientists, data engineers and researchers.
Graphical user interface. Python and Cloud Foundry’s CLI tool accommodate coding requirements.
Scripting Languages/Formats Supported
Most but through a Bring Your Own Model (BYOM) approach that requires TensorFlow.
Tight integration with other SAP products, solutions and services. APIs support document extraction, image feature extraction, image classification and other tasks, and in some cases connect to open source applications, such as Spark ML, R and Python. Provides access to cloud databases via JDBC/ODBC.
Reporting and Visualization
SAP Leonardo Machine Learning foundation stores tenant ID (identity-zone, JWT token’s zid) in application and audit logs which are sent to the SAP Cloud Platform’s ELK stack and audit log service respectively.
SAP pricing is based on node hours used within the cloud. Billing blocks are charge in increments of €7.25.
SAP Leonardo Overview and Features at a Glance:
|Vendor and features||SAP Leonardo|
|ML Focus||Broad and deep data science and ML platforms.|
|Key Features and capabilities||Fully supports a BYOM approach. Strong open source integration. Runs in SAP’s HANA public cloud.|
|User comments||Customers give SAP high marks for usability and performance. Some complain about the lack of visualization tools for external data.|
|Pricing and licensing||Tiered model based on node hours consumed in the cloud. Billing blocks are charge in increments of €7.25.|