APMdigest invited industry experts — from analysts and consultants to users and the top vendors — to predict how APM and related technologies will evolve and impact business in 2019. Part 4 covers IT Operations Analytics including Machine Learning and AI.
CONVERGENCE OF APM AND AIOPS
As machine learning and AI-related technologies continue to evolve, the boundary between APM and AIOPs will continue to soften. APM vendors that remain primarily monitoring solutions will still provide a unique and valuable service, but they will be increasingly seen as possible data sources for larger, AIOps or other advanced IT analytics platforms. On the other hand, those APM solutions that progressively invest in predictive, prescriptive and if/then analytics will have to rearchitect, not only along heuristic lines, but also to expand their reach in data sources and data types. This will create a new middle ground that currently does not have a pervasive, industry acronym.
VP of Research, Enterprise Management Associates (EMA)
AIOps will upset the traditional APM and management paradigm. As applications become more dynamic and ephemeral, the broader services that they provide now include an increasingly diverse set of technologies. These include anything from serverless and off-the-shelf components and containerization, to custom-built capabilities written in numerous languages. Application monitoring must become more infrastructure-aware in such a heterogeneous and complex environment. Future application monitoring will require a more diverse set of data collection approaches and an overarching layer of AI/ML technology to fuse data together, detect patterns, and autonomously drive automated actions. The emerging AIOps market is well positioned to fulfill these needs.
Solutions Architect, ScienceLogic
AIOPS POPULARITY GROWS
I'm forecasting a cloudy, code-based enterprise future and a containerized world, rife with noisy, data-filled virtual processes for IT operators to monitor and maintain. This means a primary focus for the future should be ensuring an organization's incident monitoring systems are capable of handling huge data streams. If not, maintaining the steady uptime and efficiency that an enterprise system requires will be impossible. If that's not enough, tools may lack the power of penetrating and analyzing data within and between containers. Therefore, I also predict a rise in artificial intelligence for IT operations (or "AIOps") implementation, as it grants the ability to see what's inside these containers to track and analyze the information generated.
Global IT Evangelist, Moogsoft
AIOps will drive automation, reducing the number of tools needed for IT management. Businesses need a solution that supports tool modernization, allows for rapid ingestion of new and scaling of existing datasets, and produces real-time insights that support automation of business services. In all likelihood, those solutions will be some variation of AIOps and will heavily rely on AI and ML for their automation and streamlining capabilities.
VP of Business Development and Alliances, ScienceLogic
Advanced analytics and artificial intelligence will continue becoming more highly focused and purpose-built for specific needs, and these capabilities will increasingly be embedded in management tools. This much-anticipated capability will simplify IT operations, improve infrastructure and application robustness, and lower overall costs. Along with this trend, AI and analytics will become embedded in high availability and disaster recovery solutions, as well as cloud service provider offerings to improve service levels. With the ability to quickly, automatically and accurately understand issues and diagnose problems across complex configurations, the reliability, and thus the availability, of critical services delivered from the cloud will vastly improve.
President & CEO, SIOS Technology
THE AIOPS HYPE CONTINUES
While AIOps has become top of mind for vendors and end users of technology products, and APM less so, the reality is that AIOps technologies are only as "intelligent" as the data provided to them. APM will remain the richest data source that aligns to the applications, business, and ultimately the users. In 2019 we will see a continued de-emphasis of APM and the increased emphasis on AIOps as we approach the peak of inflated expectations.
VP of Market Development and Insights, AppDynamics
AIOPS LIMITATIONS REVEALED
AI in IT operations will underwhelm and disappoint, largely due to unrealistically high expectations. As a results, AI applications that assist humans rather than replace them will start gaining more traction.
Head of Product, Jira Ops, Atlassian
If 2018 was the year of marketing Artificial Intelligence (AI) and Machine Learning (ML) for IT operations, then 2019 will be the year of cutting through the hype and revealing the true value of AI/ML capabilities. Much like the hype cycle we experienced with the cloud a few years ago, we're now moving past the buzzword phase and into real-world usage of AIOps (Gartner's category name for AI- and ML-assisted ops). And as organizations of all types and sizes experiment with AIOps solutions, they are slowly but surely realizing that meaningful AIOps isn't the "easy button" some thought it was. John Gentry
CTO, Virtual Instruments
Our fascination with the use of computing power to augment human decision-making has likely outgrown even the tremendous advances made in algorithmic approaches. In reality, the successful use of AI and related techniques is still limited to areas around image recognition and natural language understanding, where input/output scenarios can be reasonably constructed, and that will not change drastically in 2019. The idea that any business can “turn on AI” to become successful or more successful is preposterous, no matter how much data is being collected.
CTO, Sumo Logic
COVERGENCE OF APM AND MACHINE LEARNING
APM begins to move from monitoring to augmented intelligence. In 2019, the key capabilities for what makes a great APM solution will begin to change. Enterprise buyers will become increasingly interested in whether an APM solution utilizes machine learning, to automatically make recommendations on how to avoid issues or to optimize application performance. This capability will become important in environments that have a frequent release/change cadence and/or have hybrid cloud architecture.
Director of Technology Strategy, AppDynamics
The demand for predictive alerting and monitoring for preventative measures will become stronger than ever elevating solutions with machine learning to the top of the list as a critical criterion for product selection.
Technology Executive and Founder of the APM Strategies Group on LinkedIn.
We are entering an age where the entire soup-to-nuts of measuring user sentiment — everything from A/B testing to canary deployment — can be automated. In 2019, we'll see this play out more as companies are able to make dynamic decisions and test multiple hypothesis without administrative intervention. Gartner calls it AIOps, while Forrester favors Cognitive Operations. I see it as the future of APM.
AI is making huge strides in monitoring things like signals that are natural to humans (images/video/speech). However, in comparison the killer applications within IT have not emerged yet because no company knows yet how to prepare the "right type" of signals and the related feedback to allow for machine learning, and produce a strong meaningful application supporting IT management. Prediction: Within the next 3 years, companies will figure out the right mix of signals and feedback for machine learning and will create a breakthrough in monitoring strategies. The first to leverage this new strategy after gathering the right data will have the key advantage in the market. Ultimately, these tools will increase team efficiencies by enabling teams, which used to require experts, to operate through generalists providing considerable customer value.
VP of Products, LogicMonitor
INTEGRATION OF APM AND ANALYTICS
Integrating search and analytics engines alongside powerful APM capabilities like transaction topology will allow you to provide proactive customer support for your end-users in 2019. Having flexibility in your data analytics gives you the freedom to slice and dice the data any way you choose, uncovering new insights and making them actionable.
Director of Delivery, Correlsense
Read 2019 Application Performance Management Predictions - Part 5, covering the evolution of IT Operations Analytics and its impact on the IT team.