Michael Allen, VP & CTO EMEA at Dynatrace discusses AIOps, why it matters to enterprise IT and where it fits within the traditional IT landscape.
1. What is AIOps and why does it matter?
AIOps represents hope of helping organizations automate operations, turn data into precise answers and assist stretched IT teams who are struggling to piece together often conflicting insights from countless monitoring tools and dashboards. AIOps combines AI and continuous automation to analyses and triage monitoring data faster than humans ever could, making sense of the barrage of alerts they face due to rising IT complexity. AIOps matters as, when deployed correctly, it can help teams get out of the endless firefighting loop, enable them to eliminate false positives and identify which problems need to be prioritized to optimize the user experience – effectively allowing them to stretch their limited resources even further and focus on what matters.
According to research by Adroit, the global AIOps market is expected to grow to $237bn by 2025, demonstrating there is a substantial appetite for AIOps capabilities. However, it’s important for organizations to recognize that AIOps tools alone are not a silver bullet and must be part of an integrated approach to monitoring the IT environment.
2. Does AIOps replace existing technologies and approaches used to monitor and manage IT, or is it more of an add-on to what an organization is already doing?
Existing technologies and traditional approaches to monitoring have become largely redundant for IT teams, as they were using many tools, often in silos, with each covering only a small fraction of the overall environment. As a result, it’s impossible to monitor modern user experiences with metrics that don’t display the full context from across the entire IT stack.
AIOps offers a solution to this challenge, with the ability to provide a more holistic view across the entire IT environment by ingesting data from multiple tools and turning it into a single source of precise answers. However, AIOps doesn’t work in isolation and so shouldn’t be seen as something to replace existing approaches. Instead, AIOps should be part of an AI-powered platform, where AIOps is fully integrated with application performance, business analytics, user experience and infrastructure monitoring capabilities.
With advanced observability across the IT environment, and the business at large, AIOps solutions can pull any pertinent data and fully contextualize alerts using performance metrics. It’s important not to see AIOps as something that can quickly be implemented to replace or ‘bolt-on’ to an organization’s current approach, as both AIOps and other approaches work hand-in-hand with one another.
3. When it comes to AIOps, are we talking evolution or revolution of IT operations?
AIOps is part of a welcome revolution of IT operations, enabling teams to do more with less, but it’s important to recognize that it isn’t a complete solution in itself. The goal of AIOps is to simplify operations and deliver increased efficiency across an organization through continuous automation of problem detection, alert-triaging, and root cause analysis. The next natural step of AIOps will revolutionize IT even further by enabling the transition towards autonomous cloud operations. Here, the insights coming from AIOps solutions are used to enable cloud environments to dynamically adapt in real time to optimize performance for end-users, without the need for human intervention. However, to reach this state, a holistic approach to performance management is required, combining application performance insights and advanced observability across the cloud infrastructure with digital experience management and AIOps capabilities.
4. How would you characterize the relative value in working through historical data as opposed to working with real-time data?
Historical data certainly has its role in modern IT operations. For example, it can be beneficial for compliance and predictive analytics. However, training AIOps solutions solely with historical data brings an array of challenges. Given the dynamism of modern cloud environments, by the time AIOps solutions have learned a pattern from an analysis of historical data – the environment will have changed and so that pattern is no longer relevant.
AIOps needs to be able to adapt and deal with what’s going on in the IT environment right now, which is why real-time data is also important. This ensures the solution has up to date insights and can flag precise root-causes and instant answers to problems that arise in real-time, to help IT and business teams to optimize user experience.
5. Can AIOps offer valuable business insights at a more strategic level – beyond its strengths in optimizing hybrid IT operations through better monitoring and management?
It would be a mistake for organizations to only view AIOps through the lens of monitoring and managing the IT environment, as it can also be integral in driving better outcomes across BizDevOps teams and use cases. However, this can only be achieved by breaking down the silos between tools, to integrate AIOps with other solutions in the monitoring toolchain. This enables IT and business teams to tie together user experience insights and application performance data to identify their impact on business metrics. As a result, IT teams can become more strategic; armed with valuable insights from the entire business, they can make better-informed decisions and prioritize efforts to optimize digital services based on the wider impact on outcomes for the business, such as revenues and conversions.
As a further benefit, this also creates a single source of truth in a common language for the business, which enables IT departments to democratize access to their data. As a result, they can free up their own time by encouraging other teams to self-serve the information and insights they need to drive their own decisions, rather than needing to request it from the IT department.
6. And are there integrated, single vendor AIOps solutions available today, or is it more about acquiring two or three key pieces of software which together form the basis of an AIOps implementation?
There are many standalone, single vendor AIOps solutions available on the market, but simply adopting these tools as a bolt-on and expecting them to solve all the business’ problems is unlikely to result in success. Instead, integration with other monitoring capabilities is the secret to success when adopting AIOps. Without the other necessary monitoring tools across the IT environment, AIOps solutions won’t have the array of data needed to understand the full context of any problems and produce the results that the business needs.
The most successful approach can be found by adopting AIOps as part of an AI-powered observability platform. At the heart of this is a real time, dynamic causation model that enables true software intelligence by automatically adapting to change and negating guess work found in previous, error-prone correlation approaches. Teams are instead provided with precise answers detailing root cause, enabling them to prioritize insights based on the business outcome and impact on user experience.
7. What advice can you offer organizations who are approaching AIOps for the first time?
The key piece of advice for organizations beginning to evaluate the potential of AIOps is to understand that it is not a silver bullet to all of IT and cloud operations teams’ problems. If AIOps is simply adopted and treated as another cog in the machine, alongside an array of other monitoring solutions already in use, it will fail to reach its full potential. For AIOps to truly deliver on its promise to make life easier for IT teams, it must be part of a platform that offers a holistic approach to performance management. This integrated approach enables BizDevOps teams to not only automatically find and triage issues but create effective software intelligence that can surface answers to those problems and help to optimize the business in real-time.
Michael Allen, VP & CTO EMEA, Dynatrace