AIOps In the Post-Coronavirus New Normal: Where IBM’s Heading

Automation is central to the promise of enterprise and cloud IT operations that are truly self-monitoring, self-managing, and self-optimizing.

AIOps Keeps Data Centers in Operation Through Unforeseen Staff Shortages

Though no one truly foresaw the hit that IT staffs are taking from the current pandemic-triggered lockdowns and distancing, it’s clear that prior investments in data-center automation are lessening COVID-19’s operational impact.

AIOps refers to the use of machine learning (ML) in IT service management environments to automate more workloads that had previously relied on manual methods.

In times such as the present, AIOps tooling allows data centers to scale back their reliance on human IT administrators who may be sickened or quarantined from viral outbreaks such as COVID-19. Even when the threat of mass infection has passed, AIOps lets data center administrators mitigate operational risks that stem from labor shortages in key IT service management skills.

See also: Keeping Pace with a Maturing AIOps Market and Its Adoption

Assessing IBM’s AIOps Solutions

In the face of technical staffing challenges such as pandemics, enterprise IT professionals are increasingly relying on commercial AIOps platforms to keep their computing environments operating flawlessly.

Major IT vendors such as IBM have made AIOps a key thread for binding together their diverse platforms and tools. In fact, AI-driven IT service management is becoming a key differentiator for multicloud management tool vendors.

IBM has placed its AIOps chips behind Netcool Operations Insight, and it was a smart move. This well-established network-event reduction solution minimizes the need for manual intervention in the most common IT service management use cases.  It does this by leveraging ML to automate:

• Reduction in system noise and incidents
• Contextual correlation of events into fewer trouble tickets
• Presentation of a consolidated view across local, cloud and hybrid environments;
• Acceleration of the proactive detection of hidden issues and the identification of root causes
• Delivery of actionable insight into the performance of services and their associated dynamic network and IT infrastructures
• Provision of both real-time and historical visibility into distributed system topologies, as well as into the overall health of the network
• Execution of guided or fully automated responses for rapid issue resolution.

How IBM Should Evolve AIOps in its Multicloud Strategy

Multiclouds will only grow in importance in our lives in the wake of the COVID-19 emergency, and IBM will be missing a major opportunity if it doesn’t automate the management of those environments to the hilt. Though Netcool Operations Insight can manage practically any smart device, system, or application that produces events, it currently is limited to on-premises environments that run IBM Cloud Private.

In 2020 and beyond, IBM should consider extending Netcool Operations Insight to support automated IT service management across complex cloud topologies. Specifically, the vendor should integrate the tool with IBM Multicloud Manager, IBM Cloud Automation Manager, and IBM CloudPak for Multicloud Manager. By aligning the AIOps features of Netcool Operations Insight with IBM Cloud Automation’s Watson-powered optimization features, the vendor would be able to infuse real-time event-driven AIOps into every containerized cloud environment supported by its tooling.

IBM will need such add-on capabilities to keep in the forefront of the multicloud management arena, in which its tools face such competitors as Cisco Systems Inc.’s CloudCenter Suite and VMware’s Vrealize Operations and SD-WAN by VeloCloud.

Takeaway

Without continued improvements in the AI that drives IT service automation, no solution provider can hope to prevail for long in this segment. IBM should focus on providing AIOps tooling and services to automate the process of keeping multiclouds continuously self-monitoring, self-healing, and self-optimizing.

Going forward, IBM should align its AIOps strategy with following ongoing AI initiatives:

• Data Science Elite Team: This global group of experienced technical professionals helps IBM customers solve real problems with AI. IBM should deploy industry-focused data scientists to assist in the training of the specific AIOps assets that automate each client’s specific multicloud environment. Through this team’s agency, IBM AIOps tooling should be able to leverage pretrained AI models as well as those that learn from scratch or through training on customers’ systems and network data.
• Watson OpenScale: This solution measures, tracks, and explains AI-driven business outcomes in a single console and tunes process-embedded AI to create a feedback loop that continuously improves outcomes. It should build AIOps-focused OpenScale applications that ensure the continued alignment of AI-automated IT service management to key business performance metrics. One key AIOps feature that IBM should add to Watson OpenScale is intent-based networking, which uses embedded ML IT administrators’ intent regarding the business and technology outcomes to be achieved through automated system monitoring and management.
• Watson AutoAI: This capability – which is available within IBM Watson Studio and Watson Machine Learning – uses AI to automate the building of optimized AI models. IBM should integrate Watson AutoAI into its AIOps infrastructure to spare IT service managers from the need to manually prepare the training data or to build and train the ML models that sustain end-to-end IT service management automation.

In the growing AIOps segment, IBM’s strong suit will continue to be its world-class portfolio of AI, advanced analytics, and data management offerings, including but not limited to its Watson portfolio.

Read the other blogs in this series:

• Quantum AI in the 2020s and Beyond: What IBM Is Doing
• Business Continuity In the COVID-19 Emergency: How IBM is Responding