SAP Data Hub and SAP Data Intelligence: Streamlining Data-Driven Intelligence across the Enterprise

Part 2 of 5 in the What Is Data Intelligence blog series.

Introducing SAP Data Intelligence

SAP just released a new cloud service: SAP Data Intelligence. This blog post covers what it is and how it relates to SAP Data Hub. You can also learn about the relationship between SAP Data Intelligence and SAP Data Hub in this recording: “I just heard about SAP Data Intelligence, what is it and how does it relate to SAP Data Hub”.

SAP Data Intelligence is a new cloud service available on SAP Cloud Platform that provides the following:

  • All capabilities of SAP Data Hub, available as a service in the cloud
  • Additional machine learning capabilities:
    • Inclusion of Jupyter notebook for the ML experiment and design phase
    • Machine Learning Scenario Manager to manage all ML artifacts (e.g. models, data sets and pipelines) in one central place and also run pipelines for training or serving models

SAP Data Intelligence is the capabilities of SAP Data Hub and the SAP Leonardo Machine Learning Foundation, combined into one integrated service in SAP Cloud Platform. In the graphic below notice the machine learning services built on top of the Data Management and Orchestration capabilities provided by SAP Data Hub.

 

 

When you launch SAP Data Intelligence, you see the SAP Data Hub Launch pad with a new application called “ML Scenario Manager”.   This is where you can access the new capabilities.

 

Below is a view of the Jupyter notebook embedded within SAP Data Intelligence:

Frequent questions

So, let’s address the common questions:

  1. What does this mean for SAP Data Hub?
  2. What does this mean for Leonardo Machine Learning Foundation?
  3. What is the roadmap?

1. What does this mean for SAP Data Hub?

  • The new capabilities in SAP Data Intelligence will be available very soon in SAP Data Hub. Existing Data Hub customers will receive all new features via their maintenance agreement, at no additional cost. In other words, you can think of SAP Data Intelligence and SAP Data Hub as the same solution: the former is offered as-a-Service in SAP Cloud Platform, the latter is offered as a BYOL product that you can deploy on any of the supported Kubernetes environments (hyperscalers, private cloud or on-prem), and has a release schedule that follows the as-a-Service version by just a few months, for the new machine learning-specific capabilities.
  • Current SAP Data Hub version can already connect to Machine Learning services and APIs, including Leonardo Machine Learning Foundation (assuming you had the license), as well as open source or 3rd party ML frameworks (e.g. R, Python, Spark, TensorFlow etc.). The main difference will be that as the SAP Data Intelligence new ML capabilities are rolled-into SAP Data Hub you would no longer have to license Leonardo ML Foundation separately. Of course, you can also still use any other open-source or 3rd party ML frameworks too, both in SAP Data Hub and in SAP Data Intelligence.
  • SAP is totally committed in pursuing a multi-cloud strategy and supporting hybrid deployments: we will continue to invest to offer Enterprise AI integrated with Data Orchestration and Data Governance capabilities both as-a-Service (SAP Data Intelligence) and as a BYOL product (SAP Data Hub), to allow our customers to choose the best mix that fits their needs. Investments are equally safe on both options.

2. What does this mean for Leonardo Machine Learning Foundation?

  • The Leonardo Machine Learning Foundation evolves to become a capability of SAP Data Intelligence. Therefore, it will not be available anymore as a standalone service for new customers, who can buy it and consume it via SAP Data Intelligence.

3. What is the roadmap?

  • The SAP Data Intelligence and SAP Data Hub roadmap are the same, located here as the SAP Data Hub roadmap (you can also search for it at https://www.sap.com/products/roadmaps).
  • You will notice that all new features come to both SAP Data Hub and SAP Data Intelligence.

Key takeaways

So, what does all this mean? Here are the key takeaways:

  1. If you have SAP Data Hub today or are considering using SAP Data Hub, you should continue knowing your investment is safe. The deciding factor on SAP Data Hub or SAP Data Intelligence is the deployment type, would you rather subscribe to a cloud service, or buy a perpetual license and deploy it in a Kubernetes environment of your choice (either in the public cloud, in a private cloud or on-premise).
  2. You can use SAP Data Hub and SAP Data Intelligence for their comprehensive capabilities, which are the same:
    1. Data orchestration across complex data landscapes, which can include distributed locations, different kinds of data, different kinds of processing engines etc.
    2. Data scientist design, development and model lifecycle management with Jupyter notebook and ML Scenario Manager
    3. Metadata management and data cataloging, labeling and profiling
  3. The major use cases of SAP Data Hub and SAP Data Intelligence remain the same:
    1. Machine Learning and Data Science
    2. Business Application Transformation
    3. IoT ingestion and orchestration
    4. Data Warehousing

In summary, the top reasons to consider SAP Data Hub and SAP Data Intelligence are:

  • Scale AI across the enterprise: enable AI projects to move from POC to production and create an AI assembly line to better manage the data scientist/machine learning/AI processes, bringing IT and data scientists together.
  • Extract value from distributed data: orchestrate any data across any complex landscape, processing it close to where it lives using the most suited engine for each kind of data, minimizing the amount of data movement, and joining the dots across data silos.
  • Embrace Open Technologies: SAP Data Intelligence and SAP Data Hub both execute in Docker containers, orchestrated by Kubernetes. Both leverage open source heavily, and natively support many ML frameworks, both SAP and open-source or 3rd party, such as TensorFlow. They also can execute R, Python, Spark, and other open-source frameworks and are agnostic to the underlying infrastructure (as long as a supported Kubernetes distribution is available).

This article is part 2 of 5 in the What is Data Intelligence blog series; stay tuned for the rest of the posts—coming soon.

 

Related blog posts:

VN:F [1.9.22_1171]
Average User Rating
Rating: 5.0/5 (8 votes cast)
SAP Data Hub and SAP Data Intelligence: Streamlining Data-Driven Intelligence across the Enterprise, 5.0 out of 5 based on 8 ratings

2335 Views