Top AIOps trends to watch out for in 2020

Profile picture for user Guy Fighel By Guy Fighel January 14, 2020
Summary:
Amid the noise of information overload, AIOps is set to grow as companies prioritize digital transformation. Guy Fighel of New Relic shares top trends for the coming year

Code with brain artificial intelligence AIOps concept © Antonov Serg - shutterstock

As software systems find themselves overwhelmed by incessant flood of information, the demand for accurate troubleshooting and incident resolution efforts are emerging as a top priority for IT teams. Between noisy alerts, signals distributed among multiple tools, and thousands of ‘unknown unknowns’, it’s becoming increasingly difficult for IT teams to quickly determine and address the root cause of incidents, let alone detect and respond to issues proactively.

AIOps – the use of big data and machine learning to automate IT operations – helps simplify this alert fatigue, and we will see its growth accelerate as companies make digital transformation a key business priority. In fact, Gartner, who first coined the term, forecast that large enterprise exclusive use of AIOps and other tools will grow from 5% to 30% between 2018 and 2023.

With AIOps slowly gaining momentum among IT teams, here are the top six trends to watch out for this year.

IT leaders will truly take note of AIOps

Forecasts by IDC predict that worldwide spend on AI systems will reach $77.6 billion (about £58.7bn) in 2022, over three times what it had estimated for 2018. Whilst AI has an extraordinary number of use cases, this year, IT leaders will truly take note of AI’s usefulness for IT operations as they undergo digital transformation. The C-suite is waking up to the short- and long-term benefits of AIOps, reducing downtime and significantly boosting bottom lines, which can only be a good thing, as the lives of service desk, DevOps and InfoSec teams will all get easier as investment into AIOps grows. Furthermore, as AIOps tools mature, they will be able to process a wider variety of data types and deliver value faster and better, enhancing performance for more specific tasks.

Incident management functions will be augmented

AIOps will be used to augment natural language processing, root cause analysis, anomaly detection, event correlation and analysis, as well as other IT functions to offer IT operations professionals greater control. Event correlations and incident intelligence will be seamlessly available within an active teams’ ideal incident management platform. Plus, anomalies will be pre-emptively detected and highlighted within the collaborative tools organisations already use to maximum effect.

The emergence of data-agnostic AIOps tools

In 2020, vendors are expected to launch data-source-agnostic tools. These new solutions will be a big win for the industry as the more varied the data that is ingested into an AIOps platform, the greater the value (insights) the algorithms can accomplish. The more ‘openness’ (as it is sometimes called) that there is, the more the re-engineering of the core data platform there can be to take in a wide range of data sources and types of data to drive deeper visibility. The goal is to build fantastic visibility of the context to relate data such as events, metrics, logs and traces to one another. This means users can accurately determine issues, predict impacts and understand how change can affect business-critical activities. Knowledge is power.

AIOps becomes an embedded capability of observability

At the moment, the AIOps solution landscape is quite fragmented. Vendor point solutions and ancillary technology providers are converging, but over the next 12 months there will be an increasing expectation for AIOps to be an embedded capability of observability. Observability platforms that enable IT operations teams to leverage all of the telemetry and operational data they need for AI and machine learning will be key to speedier incident response.

Automatic remediation will take to the stage

As DevOps teams work to bridge the gap between detecting incidents, diagnosing issues and remediating problems, the scope of AIOps will enlarge to include more automatic remediation. This will mark a clear step change for IT companies. This is because thus far, most AIOps capabilities have centered on DevOps teams themselves using tools for detecting, diagnosing and prioritizing – proactively detecting anomalies, accelerating and improving understanding through event correlation and incident intelligence, and then affording aid to recommend how to act and escalate difficulties to the best individual or team.

DevOps teams will get up to speed with AIOps tools faster

There are still far too many AIOps tools on the market that demand a lot of time setting up and implementing before they generate real value. Time-poor DevOps teams are being held back by the long onboarding times these tools require, as well as the time it takes importing data into them, and also educating and training their colleagues on how they work. If AIOps is to reach the masses, which it will, this challenge cannot and should not continue. Fortunately, in 2020, we can expect AIOps capabilities to leverage a flexible baseline of rules that change based on the input of the user and production data over time. DevOps professionals will be able to insert their own logic into the system, in addition to accepting or providing feedback on the logic that is automatically proposed.

This year AIOps tools will become simpler to realize, be taught and used, and the ones that do not will become obsolete. The days of dedicating hours upon hours to develop and train models for each new integration point are numbered.