Virtualization Technology News and Information
Article
RSS
VMblog Expert Interview: Masergy Explores Results from their State of AIOps Study

interview-masergy-pandya 

Masergy is an early proponent of AIOps, using it as a virtual network assistant and to automate network operations in the push toward autonomous networking. 

Looking into AIOps, the company recently announced its new State of AIOps study, conducted by ZK Research.  In it, a resounding majority (86%) of IT leaders predict that they will have a fully automated network within 5 years.

To find out more interesting information like this, VMblog reached out to Ajay Pandya, director of product management at Masergy.

VMblog:  Tell me about your latest AIOps research study and what makes it particularly relevant for today's CISOs, CIOs, CTOs, etc.

Ajay Pandya:  Today's IT organizations are particularly overwhelmed, especially with a surge in remote work and hybrid business models making more demands on them -- from the network and security to scarce time and resources. In response, many IT leaders are turning to AI for IT Operations, or AIOps, to address those issues. AIOps takes machine learning, behavioral analytics, and predictive analytics and applies them specifically to the challenges of IT. It can make operations smarter, giving teams a more intelligent view of their network and application performance as well as boosting their security posture. In designing our AIOps study with ZK Research, we assumed that most IT leaders understand that AI is beneficial, but we wanted to dig deeper into the adoption drivers and benefits. 

VMblog:  What was the most surprising research finding about the State of AIOps?

Pandya:  IT leaders are overwhelmingly embracing AIOps. The study revealed an incredible amount of excitement, trust, and confidence in the technology. 64% of IT leaders report that they are already using AIOps and more than half (55%) are using it for both network performance and security, gaining more holistic insights. It's clear that IT leaders no longer see AI as experimental but rather as a necessary competitive advantage and a vital path forward in reducing IT management headaches.

VMblog:  What are the top use cases and what investments are needed to gear-up for AIOps adoption?

Pandya:  Survey respondents identified the top use cases for respondents as cloud app analytics, performance improvement, network service optimization, and faster threat detection and response. Sixty-four percent of the respondents measure AIOps success based on IT operational efficiencies, and another 54 percent on improved network or app performance. These results point to tangible benefits of AIOps improving business operations. The study also found that network teams spend nearly 50% of their time monitoring application performance and network troubleshooting, which further explains why AIOps efficiency gains are a much-needed relief for most IT organizations.

Readiness is important, according to the findings. IT leaders are preparing for AIOps technology by investing in:

  • Software-defined network modernization and virtualization (73%)
  • Cloud migration (67%)
  • Training for AI models (61%) 

Indeed, IT infrastructure matters, and legacy networks aren't designed for AIOps. AIOps is only as smart as the data it is fed, and siloed data doesn't compute. Networks built on software defined principles, SD-WAN, and SASE architectures have centralized management tools and unified orchestration systems that allow for AIOps to gather and evaluate data across both the network and security. Among survey respondents, 77% agree that AIOps performs better with a SASE architecture, where SD-WAN and network security tools are all in one platform. In fact, this is the top-ranked consideration for employing AIOps.

VMblog:  Is there an endgame for AIOps or like other technology paradigms, does the goal post keep moving?

Pandya:  The end game is full automation or a fully autonomous network, where IT systems are self-managing and self-healing without human intervention. A resounding majority (84%) of IT leaders see AIOps as the path to a fully automated network environment. Eighty-six percent predict that they will have a fully automated network within 5 years, and 97% already express a high level of trust in AIOps tools to act on their own recommendations and create fully automated systems.

Clearly, AIOps is the key technology in building an autonomous network, and while some IT leaders may believe they are using AIOps solutions, not all solutions are equal. 

There's a big difference between using AI to gain insights and using AI to power automation. In order for an AIOps solution to help companies reach the milestone of full autonomy, it must be given access to IT control panels, where it can automatically make adjustments to network configurations and security policies. Without these permissions, AIOps will be nothing more than an alerts generator, making IT staffers do the actual work of responding to its recommendations. And machine learning is key -- AIOps should get smarter with time and training. For these reasons, due diligence is critical.

VMblog:  What does the journey look like to reach the horizon? Is it further away or closer than we think?

Pandya:  The study findings reveal the issues companies face, which in turn can be used as lessons learned in smoothing the path toward AIOps. Respondents state their biggest AIOps challenges as:

  • Eliminating system fragmentation (50%)
  • Staffing support (47%)
  • Training for AI models (47%)

Staffing and training for AIOps are equal and related challenges highlighting the importance of expertise and playbooks that should guide any AIOps program. Reaching the milestone of full autonomy doesn't happen overnight, and can be accelerated with the help of a fully managed service. IT organizations must train the AIOps engine on their operational standards so they can learn to trust the machine to act alone. This is ultimately how to evolve through key stages of automation and arrive at an autonomous network. 

Four progressive steps are part of the journey toward autonomous networking:
  1. AIOps Insights: "Here's the problem I found."
  2. AIOps Recommendations: "Here's the problem and my recommended repair."
  3. AIOps Partial Automation: "Here's my recommendation. Approve and I will make the repair."
  4. AIOps Autonomous Networks: "Here's what I just fixed for you."

The end game of autonomous networking is closer than you might think. Afterall, everything IT teams need to build a fully autonomous network is already available today. It's only a matter of implementing the right technology and service provider and then stepping through the process of training the machine and building trust.

VMblog:  Where can VMblog readers go if they'd like to download the report or learn more?

Readers can view the full report here: https://files.masergy.com/hubfs/White%20Papers/2021%20State%20of%20AIOps%20Study.pdf

##

Published Thursday, December 16, 2021 7:32 AM by David Marshall
Filed under: ,
Comments
There are no comments for this post.
To post a comment, you must be a registered user. Registration is free and easy! Sign up now!
Calendar
<December 2021>
SuMoTuWeThFrSa
2829301234
567891011
12131415161718
19202122232425
2627282930311
2345678