See where you land on 451 Research's spectrum of AI use cases for data management

Portfolio Marketing Manager, Hybrid Data Management, IBM

68 percent of surveyed businesses recently responded that they use machine learning (ML) or plan to do so in the next three years. AI technologies rapidly are becoming how businesses distinguish themselves from competitors. But choosing the best way to implement AI isn’t always a straightforward process.

The AI use case spectrum for data management featured in 451 Research’s recent survey and pathfinder report, Accelerating AI with Data Management; Accelerating Data Management with AI can help guide businesses exploring AI. From automation to augmentation, 451 Research’s recommendations can help put your business on pace to join the 92 percent of respondents with positive opinions of ML and AI project performance.

Automating the simple and the complex

Automation use cases live at both ends of 451 Research’s spectrum. On one side there are extremely repetitive tasks that don’t tap into the full breadth of a worker’s skills and experience. The other side shows extremely complex tasks that would take a person far too much time and effort to complete—if even possible at all. Both of these situations are prime candidates for automation so that employees can put their time and effort into activities where they can provide the most value.

For example, take a look at manual data preparation.

39 percent of respondents believe data ingestion and preparation is the most demanding stage of the AI process. It can be frustrating for skilled workers to spend a significant amount of time on these tasks as opposed to actually implementing the data they intend to use.

Automating tasks such as data identification and tagging through machine learning can free up employees’ time so that they can easily engage in solving more intricate, high-value challenges.

Looking for efficiencies in running queries also falls into this category. You need to ensure that queries are optimized to take the best path and are managed for stable and reliable resource consumption. This can help you obtain insights and act on them more quickly, giving you greater potential to make decisions based on those insights when they are most valuable and haven’t been discovered by competitors.

Data exploration is a good example of a complex task that benefits from data management. If you can accelerate the exploration with the right AI-infused database tools, you can develop AI applications much more efficiently than if that exploration needed to have been completed manually. This saves time and allows your teams to place their efforts where they add the greatest value.

Augmenting employees’ ability to do everything else

Everything that cannot be automated falls in-between the two ends of the spectrum and should use AI to augment the capabilities of existing personnel. 451 Research highlights two roles in particular that will benefit greatly from doing so: business analysts and database administrators (DBAs).

The relative shortage of data scientists – estimated by 451 Research to be between one and two million – will need to be bolstered by the estimated 55 to 75 million business intelligence users and 200 to 250 million knowledge workers. Data management offerings built for AI can shift some of the workload to knowledge workers in several ways. When data scientists can develop AI applications more quickly, business analysts can use the insights generated by them to make decisions as opposed to raising multiple, ad hoc requests. Natural language querying also improves business analysts’ ability to find answers they may not have the coding knowledge to access in traditional systems. A self-service model can save everyone time.

451 Research indicates that DBAs benefit from AI to partially automate tasks such as:

  • Database provisioning and patching
  • Performance tuning
  • Backups, high availability and disaster recovery
  • Schema changes

While companies will likely not want to completely entrust their business to AI functionality in all of these areas quite yet, using it to help DBAs ability to perform their job more efficiently can be beneficial to your business . AI can propose opportunities or take actions that are allowed or rejected based on DBA oversight.

Advice for businesses considering an AI use case for data management

Given the potential for AI use cases related to data management, 451 Research also provides some recommendations to help companies seeking to get started with or draw the most out of their AI implementation. Foremost, they suggest identifying potential AI use cases quickly to stay competitive with  businesses that have already found success with their AI initiatives. 451 Research also stresses the importance of early wins to ensure widespread buy-in across stakeholders and users. Simpler projects such as automation of routine tasks with AI are a great way to show that AI is a worthwhile project to champion. From there, larger scale projects with potentially more impact can be undertaken.

A final key point from the report: You should embed AI at the data level. Data is the foundation upon which the journey to AI and insight begins. Embedding AI at that level helps ensure you can carry benefits through the data management lifecycle and companies as a whole. From ingestion to driving applications, implementing AI in its initial stages is key.

If your business is ready to bring AI in at the data management level, look into the Db2 family of data management solutions. We have infused Db2 offerings with AI and are can jumpstart your most promising AI use cases.

To learn more about data management, AI in general, or Db2 11.5, the AI database, check out: