Key Takeaways
|
|
AIOps platforms fall into two main categories: domain-centric and domain-agnostic solutions.
What are the differences between domain-centric and domain-agnostic AIOps, and why should you choose one type of solution or the other? Read on for guidance on understanding the respective pros and cons of domain-centric and domain-agnostic AIOps.
Definition of Domain-Centric AIOps
Domain-centric AIOps is the application of AIOps to solve problems within a certain domain.
For example, a domain-centric AIOps platform might help teams manage just application performance, or just network performance.
Similarly, the tool will be able to collect and interpret data associated only with a specific domain. A network-management AIOps tool is not going to collect metrics beyond those related to the network, for example.
Definition of Domain-Agnostic AIOps
Domain-agnostic AIOps is the use of AIOps to work with any type of data and solve problems in any domain using data from across the IT environment.
A domain-agnostic AIOps platform, then, might be designed to help IT teams manage all aspects of their operations, from application deployment and performance management, to network tuning, to capacity planning and cost management.
Differences Between Domain-Centric and Domain-Agnostic AIOps
The difference between domain-centric and domain-agnostic AIOps tools, then, is really pretty simple: it’s a matter of how broad the set of target use cases is.
By extension, the difference also includes the level of depth that the AIOps solution goes into. A domain-centric AIOps tool designed for network management will likely offer more features tailored to network management than will a domain-agnostic tool that can be used for network management, but is not designed specifically or solely for that use case.
It’s worth noting that, in general, all AIOps solutions target IT-related use cases. Although AIOps theoretically could be applied to address some needs within domains outside of IT, like customer relations management, use cases such as this have not yet become a focus for AIOps vendors.
So, all of the “domains” at play in AIOps are IT domains. The difference between domain-centric and domain-agnostic platforms lies in whether they focus on just one domain within IT, or all of them.
How to Choose Between Domain-Centric and Domain-Agnostic
Both categories of AIOps platforms are valuable. It’s not the case that domain-agnostic solutions are inherently better because they cover a broader set of use cases, or that domain-centric tools are better because they offer more features within a certain domain.
Instead of defaulting to one type of solution or the other, then, businesses should assess their specific requirements. Ask yourself:
- Could some of my ITSM processes benefit from AIOps more than others? If so, you may want domain-centric solutions that focus on the relevant domains. On the other hand, if your entire IT operation is sorely in need of automation, a domain-agnostic platform may help you get there.
- How many tools can my team learn and manage? Domain-agnostic AIOps may be more attractive because it doesn’t require your engineers to learn and work with multiple platforms.
- How complex are my use cases? Domain-centric AIOps is most valuable when you have truly complex use cases. For example, if you need to troubleshoot complex, microservices-based applications, domain-centric APM tools that cater to that need will do it best. But if you just need standard management features, a domain-agnostic AIOps solution will likely suffice.
- What is my budget? Overall, a domain-agnostic AIOps platform that can handle all of your use cases is likely to provide a lower total cost of ownership compared to deploying multiple domain-centric tools to address different needs.
- How will my AIOps needs change over time? If your management needs are unpredictable, or you expect that you will want to apply AIOps to additional domains in the future, a domain-agnostic platform provides the most flexibility.
Why AIOps May Become More Domain-Agnostic Over Time
Although domain-centric AIOps platforms are not likely to disappear anytime soon, the overarching trend in the AIOps market appears to be toward domain-agnostic solutions.
Circa five years ago, when AIOps was new, it made sense for vendors to create AIOps solutions that catered to specific, limited use cases. They were simpler to build – not to mention easier to adopt for businesses that felt daunted by the task of applying AIOps to their entire IT operation at once.
But now that AIOps has matured and businesses are looking to apply it to multiple needs across their IT estates, domain-agnostic tools are in a position to prove more attractive overall.
Again, that’s not to say domain-centric tools are becoming obsolete – they are not – but going forward, expect to see more and more focus on domain-agnostic solutions.
View the Gartner Market Guide for AIOps Platforms to learn more about the AIOps market.
Tag(s):
AIOps
Chris Tozzi
Chris Tozzi has worked as a journalist and Linux systems administrator. He has particular interests in open source, agile infrastructure, and networking. He is Senior Editor of content and a DevOps Analyst at Fixate IO. His latest book, For Fun and Profit: A History of the Free and Open Source Software Revolution, was...
Other posts you might be interested in
Explore the Catalog
October 4, 2024
Capturing a Complete Topology for AIOps
Read More
October 4, 2024
Fantastic Universes and How to Use Them
Read More
September 26, 2024
DX App Synthetic Monitor (ASM): Introducing Synthetic Operator for Kubernetes
Read More
September 16, 2024
Streamline Your Maintenance Modes: Automate DX UIM with UIMAPI
Read More
September 16, 2024
Introducing The eBPF Agent: A New, No-Code Approach for Cloud-Native Observability
Read More
September 6, 2024
CrowdStrike: Are Regulations Failing to Ensure Continuity of Essential Services?
Read More
August 28, 2024
Monitoring the Monitor: Achieving High Availability in DX Unified Infrastructure Management
Read More
August 27, 2024
Topology for Incident Causation and Machine Learning within AIOps
Read More
August 23, 2024