The IT operations landscape is changing rapidly with fast adoption of dynamic infrastructures, including hybrid clouds, containers, and microservices. Legacy monitoring vendors are struggling to adapt their products to fit these modern architectures and to deal with the volume, variety and velocity of data generated. Old delineations between monitoring tools (such as infrastructure versus network versus application) are breaking down. The glue that brings the layers together is innovations in the application of Artificial Intelligence (AI) for IT Operations.
The AIOps Maturity Model will help organizations to understand their AIOps Maturity and high-level strategies to grow their AIOps capabilities.
Perceive - Reason - Act Paradigm
The AIOps maturity model is based on the perceive-reason-act paradigm. The perceive-reason-act cycle in AI is the connection between the system and the environment. This cognitive feedback loop consists of a perception phase, where the state of the system and the environment is observed, a reasoning phase where the perceived state is analyzed, and an action phase where an action is executed. In short, acting requires reasoning and reasoning requires perception. There are several important ways this dependency manifests itself in the AIOps space and guides the strategy of AIOps adoption.
The value of AIOps
The main value to the company lies along the ACT dimension – perception and reasoning are as useful as the actions they enable. Therefore, the main aspiration for a company should be advancing along the ACT dimension. In order to advance along the ACT dimension, a corporation has to mature its perception and reasoning capabilities – these two efforts can be quite disconnected; therefore, it is important to have a good strategy of how to advance them. There are three main levels in the ACT dimension:
- No automation
- Human Automate Ops tasks
- AI Automates Ops tasks
AIOps Growth Strategies
There are three strategies to advancing along the ACT dimension:
- Specialist: the specialist approaches the AIOps tasks like anomaly detection, noise reduction in alerts or KPI predictions as separate independent tasks.
- Generalist: the generalist aims to get full visibility of the IT landscape and business in order to approach AIOps tasks holistically.
- Hybrid: With this approach an organization attempts to balance between the Generalist and Specialist strategies with the goal of taking the best of both worlds.