In APMdigest's exclusive interview, Colin Fletcher, Research Director at Gartner, talks about Algorithmic IT Operations (AIOps) and the challenges and recommendations for AIOps adoption.
Start with Gartner Talks About AIOps - Part 1
APM: What are current tools lacking that has given rise to AIOps?
CF: Many tools used across IT Operations Management historically and even today are implicitly or explicitly built around paradigms that limit their ability to cope with or keep up with rapidly changing demands and technological innovation.
Examples of these design paradigms include but are not limited to:
■ A rigid linkage between the data collection mechanism, the data retention mechanism, the analytical engine(s), and the means to visualize and interact with their results
■ Analytical capabilities that depend heavily on manual development and maintenance to deliver value
■ Data ingestion and storage mechanisms that do not easily scale
■ Limited support of bidirectional, "full" integration with other tools
■ Best practice-formed workflows and processes that are rarely adapted, updated or otherwise modified
These paradigms evolved for good reason out of decades of operational experience, however these same paradigms are exactly the calcified barriers to meeting growing and diversifying needs of business stakeholders.
APM: Is this tool used on top of a company's existing analytics and other IT Operations tools, or does it replace them?
CF: Currently most organizations are using AIOps platforms in tandem with other IT operations management tools and we expect that will continue to be the case for some time. That said, more and more enterprises and in particular DevOps teams find that over time their AIOps platform should be considered to be used in place of multiple tool types, particularly legacy tools.
APM: Is a certain type or size of enterprise a better candidate for AIOps?
CF: Keeping in mind that AIOps encompasses a wide span of capabilities, from basic log/machine data analysis tools to sophisticated, machine learning analytical engines, we see enterprises of all types finding good places to start on their AIOps journey.
APM: Does AIOps require a new skillset in IT?
CF: As mentioned earlier, AIOps covers a wide span of tooling, so it's possible to get started on an AIOps journey using very basic skills to start analyzing logs for example. That said, taking maximum advantage of all that AIOps has to offer will require an investment in developing or acquiring a level of analytical skill not always found in IT operations teams. Enterprises often find that DevOps teams/initiatives and business analysts are often good internal sources of talent.
APM: What do you see as the biggest barriers to AIOps adoption?
CF: Misplaced fear that the automated analytical capabilities AIOps offers will directly lead to job reductions, real and perceived cost issues, previous negative experiences with prior generations of statistical pattern discovery and recognition and/or event correlation and analysis tools that failed to deliver, fear that AIOps tools require an unobtainable level of skills to be useful, "tool gravity"/reluctance to change.
As you probably detected, most of these issues while reasonably formed, are rooted in perceptions built around a literal previous generation of thinking and technologies that does not directly apply to today's AIOps tools.
APM: For enterprises starting out with AIOps, do you have a recommendation of where to start?
CF: AIOps tools have progressed to the point that many of them are actually very easy to just look at and try with little or no cost – I think that's a great way to get a sense of what these tools are capable of, and I wouldn't wait long to do so! Of course Gartner clients are always welcome to also take advantage of our published research on the topic or get some time on the calendar to talk about how AIOps can help them.
ABOUT Colin Fletcher
Colin Fletcher focuses his research on how advances in application release automation (ARA), IT operations analytics (ITOA), continuous configuration automation (CCA) and DevOps can help IT operations teams continually drive greater business success, reduce costs and mitigate risk. Fletcher's research is informed by daily conversations with clients and thought leaders, as well as more than 16 years of IT practitioner experience (from service desk to admin to consultant), leadership experience (team, project and product management) and creative marketing experience (product and strategic) built at companies large and small, including Apple, HP, BMC Software, Motorola, IBM Global Services, Dell and several startups.
The development of the Thousand Brains Theory of Intelligence framework will now serve as a foundation for further research and new developments in Artificial Intelligence (AI) and Machine Learning (ML) ...
IT teams feel overwhelmed by too many tools that do not provide a unified view of the entire IT infrastructure, according to The Shift to Unified Observability: Reasons, Requirements, and Returns, a new independent survey conducted by IDC in collaboration with Riverbed ...
Legacy systems require a great deal of a prior knowledge, and then significant configuration, for anomaly detection to work effectively. ML and AI are beginning to change that, but it's important to really validate the claims of any NPM solution ...
Successful insight into the performance of a company's networks starts with effective network performance management (NPM) tools. However, with the plethora of options it can be overwhelming for IT teams to choose the right one. Here are 10 essential questions to ask before selecting an NPM tool ...
Hybrid and remote work environments have been growing significantly in the past few years. As individuals move away from traditional office settings in today's new remote and hybrid environments, many operational issues such as poor visibility into asset status and refreshes, unaccounted assets, and overspending on software are becoming a bigger challenge for IT departments ...
MLOps or Machine Learning Operations are a combination of best processes and practices that businesses use to run AI successfully ... While it is a relatively new field, MLOps is a collective effort that captured the interest of data scientists, DevOps engineers, AI enthusiasts, and IT ...
The data is in: enterprises are not happy with their managed service providers (MSPs) and cloud service providers (CSPs). According to the latest CloudBolt Industry Insights report, Filling the Gap: Service Providers' Increasingly Important Role in Multi-Cloud Success, 80% are so unsatisfied with their existing MSP and/or CSP, they are actively looking to replace them within 12 months ...
The last two years have accelerated massive changes in how we work, do business, and engage with customers. According to Pega research, nearly three out of four employees (71%) feel their job complexity continues to rise as customer demands increase, and employees at all levels feel overloaded with information, systems, and processes that make it difficult to adapt to these new challenges and meet their customers' growing needs ...
Investing in employees will always be smart business. And right now, investing in employees means giving people the resources — and ability — to optimize performance ... For pretty much every company, that means delivering the digital tools necessary to facilitate seamless, secure, user-friendly access and connectivity ...
Digital transformation can be the difference between becoming the next Netflix and becoming the next Blockbuster Video. With corporate survival on the line, "digital transformation" is no longer merely an impressive buzzword to throw around in boardrooms. It's the ticket for entry into the digital era, a fundamental business strategy for every modern company ...