Media Query Source: Part 58 - Database Trends & Applications (US digital magazine); Data experts and leaders for opinions on transforming data & analytics in the cloud

• Database Trends & Applications (US digital magazine)
• Data experts and leaders for opinions on transforming data & analytics in the cloud
• Public clouds support all data and analytics functions
• Considerations include use cases, construction & administration

The query responses I provided to Database Trends & Applications on August 27, 2025:

Database Trends & Applications: This special report, to run in the October issue of Database Trends & Applications (www.dbta.com), will cover how cloud services are supporting data and analytics workloads and initiatives. I am seeking the views of industry experts and leaders on how the cloud has changed the data landscape, and what data managers need to know about their choices.

Are most, if not all, data and analytics functions now supported on cloud, or are there cases where they are better left on-premises?

Gfesser: All data and analytics functions are supported via public clouds. If a desired function isn't available via a native cloud service, all of the primary public clouds offer marketplaces from which functions can be obtained via third party vendors, and custom software can also be hosted if needs aren't being met, often in combination with some level of open source software.

Cost and security are typically two primary areas of concern when it comes to use of public clouds for data and analytics functions. With respect to cost, it's extremely important to understand what one's production use cases will actually look like, who will be executing these use cases, and how often these will be executed. If you're not careful, baseline cost can be prohibitive in the cloud, but cost can also be prohibitive on premises due to upfront hardware procurement. With respect to security, AWS is known for its shared responsibility model in which AWS is responsible for security of the cloud and its customers are responsible for security in the cloud. But as with the other public clouds, it's arguably best to rely on their security of the cloud rather than taking it upon oneself to try to replicate the security level they provide out of the box. Security in the cloud is always going to be the responsibility of the customer, regardless of which of the public cloud and on premises options are chosen. All of this said, the presence of any regulations or legal restrictions should be reviewed prior to hosting data and analytics functions on public clouds.

Are there any issues or complications of which data managers need to be aware as they deploy on cloud? And is the cost of long-term cloud a barrier?

Gfesser: As mentioned previously, use cases need to be understood and weighted against costs. For example, it arguably doesn't make sense for data and analytics functions to be continuously running in the background unnecessarily due to infrequent usage, in which case serverless options provided by public clouds might make more sense due to these running only on an on-demand basis. Also important is the need to assess the talent who is building and administering data and analytics functions, to help ensure that they are cost conscious, as it's not uncommon for inexperienced individuals to unknowingly run up costs due to cloud services running unnecessarily when not being used. In such cases, automation can help to mitigate such risks.