Apache Airflow vs. SAP Data Intelligence

For a quick job scanning of status and deep-diving into job issues, details, and flows, AirFlow does a good job. No fuss, no muss. The low learning curve as the UI is very straightforward, and navigating it will be familiar after spending some time using it. Our requirements are pretty simple. Job scheduler, workflows, and monitoring. The jobs we run are >100, but still is a lot to review and troubleshoot when jobs don't run. So when managing large jobs, AirFlow dated UI can be a bit of a drawback.

Incentivized

If you have an SAP products ecosystem in your IT landscape, it becomes a no-brainer to go ahead with an SAP Data Intelligence product for your data orchestration, data management, and advanced data analytics needs, such as data preparation for your AI/ML processes. It provides a seamless integration with other SAP products.

Incentivized

• In charge of the ETL processes.
• As there is no incoming or outgoing data, we may handle the scheduling of tasks as code and avoid the requirement for monitoring.

Incentivized

• It integrates well with our current ecosystem of SAP products, like HANA.
• It provides end-to-end machine learning operations, with tools for the complete model life cycle.
• It has a simple user interface for novice users, with complex tools also available for power users.
• It builds on SAP Data Hub, providing all the ETL functions of that tool with additional machine learning functionality.
• It can run in the cloud, no on-premise software management needed.
• Many programming languages are supported, it provides a sandbox environment for the user to develop in whichever style they prefer.
• SAP is very actively developing and improving it.

Incentivized

• they should bring in some time based scheduling too not only event based
• they do not store the metadata due to which we are not able to analyze the workflows
• they only support python as of now for scripted pipeline writing

Incentivized

• Data transfer speed tends to be slow when there is poor internet connection since SAP Data Intelligence don’t synchronize data while offline. However, this is not vendor fault, that’s why we have implemented robust wireless technology internet connection in our organization.

Incentivized

Allow collaborations among various personas
with insights as ratings and comments on the
datasets Reuse knowledges on the datasets for new users Next-Gen Data Management and Artificial Intelligence

Incentivized

Easy to learn Easy to use Robust workflow orchestration framework Good in dependent job management

Incentivized

I think the troubleshooting process might be streamlined with improved error recording and tracing. A lot of information about issues and how to fix them is hidden away in the Kubernetes pods themselves. I'm not sure whether SAP Data Intelligence can fix this problem it may be connected to Kubernetes's design, in which case fixing it could need modifications inside Kubernetes itself.

Incentivized

Initially we struggle to get help from SAP but then dedicated Dev angel was assigned to us and that simplify the overall support scenario. There is still room of improvement in documentation around SAP Data intelligence. We struggle a lot to initially understand the feature and required help around performance improvement area,

Incentivized

There are a number of reasons to choose Apache Airflow over other similar platforms- Integrations—ready-to-use operators allow you to integrate Airflow with cloud platforms (Google, AWS, Azure, etc) Apache Airflow helps with backups and other DevOps tasks, such as submitting a Spark job and storing the resulting data on a Hadoop cluster It has machine learning model training, such as triggering a Sage maker job.

Incentivized

One of the reasons to pick SAP Data Intelligence is the speed and security it provides, in addition to the excellent support it provides. It is also compatible with all popular databases, which is another reason to choose it.

Incentivized

• A lot of helpful features out-of-the-box, such as the DAG visualizations and task trees
• Allowed us to implement complex data pipelines easily and at a relatively low cost

Incentivized

• Automation of data management slashed tasks by over 60% in most departments for the first 8 months.
• Metadata catalogs have enabled us to categorize data from disjointed sources in one place.
• It runs multiple ML models which enhances flexibility when managing data.

Incentivized