Home
Case Studies
IBM Software Support and Site Reliability Engineering
Support services are the backbone of any enterprise, ensuring seamless operations, maximizing uptime and delivering exceptional customer experiences. In today’s intricate digital ecosystem, robust support services are crucial for businesses to thrive. By providing timely assistance, resolving issues efficiently, and proactively addressing potential problems, support teams play a pivotal role in maintaining business continuity, enhancing customer satisfaction and driving innovation. It is critical for these experts to spend their time on the most critical issues and tasks.
The IBM Software Support team encountered a similar challenge. Tasked with managing and resolving technical issues across an extensive portfolio of software products, the team handles hundreds of thousands of cases annually. With thousands of support engineers distributed globally, it was critical they were providing a world class support experience by being highly productive. By harnessing the power of IBM watsonx™, an enterprise-grade AI studio that helps AI builders innovate with API’s, tools, models and runtimes to build AI solutions, IBM Software Support leveraged the power of generative AI (gen AI) tools to tackle various use cases within the support landscape to help transform the way support is delivered.
The IBM Software Support team handles complex technology issues across the client stack. Productivity of this team is critical to ensure they are spending time on the right issues to provide world class client experience. Though problem resolution can be human-intensive, human skills can be augmented. That’s why they brought watsonx into the process to help increase productivity, optimize issue deflection and reduce time to resolution.
The impact across the support organization was significant. First, they architected and designed AI solutions that would address a large segment of support workload that could be automated.
The automation identifies solutions for client cases that are addressable with various IBM technical assets, including IBM product documentation and technotes. With a single client interaction, generative answers are identified that contain the information from IBM documentation or technotes to resolve a question or problem.
AI built on watsonx supercharges support efficiency by automating routine tasks, streamlining diagnostics and prioritizing critical cases. By automating responses and translations, AI built on IBM® watsonx.ai™ freed up engineers to tackle complex issues. Additionally, Analyze SW Support’s ability to analyze vast datasets and identify patterns enabled faster problem resolution.
This allows support engineers to focus with surgical precision on their specialized tasks and excel at what they do best. They’ve identified several instances where AI excels at improving efficiency to streamline the support processes, improve response times and provide outstanding service to our customers.
From resolving routine cases to identifying root causes, automating tasks and predicting escalations, IBM Software Support was able to harness the power of gen AI to develop a number of innovative support use cases. Top 3 use cases in term of results:
Ask SW Support is a RAG solution (Retrieval Augmented Generation) that leverages the power of Large Language Models (LLMs) and watsonx.ai. This cutting-edge tool automates low-complexity support cases, freeing up valuable time for SW support engineers to tackle intricate, high-priority issues. In the 4 months since this solution has been used by IBM Support engineers, we have seen an average of 10 minutes of time saved per case. With an estimated 370,000 low-complexity cases handled annually, and estimating AI-powered answers for 100,000 of them, this could potentially save the team over 80,000 minutes each month, resulting in a significant annual time savings of approximately 17,000 hours (or 1,400 hours per month).1
Ask SW Support isn't the only tool in the arsenal. Analyze SW Support brings the power of anomaly detection to the diagnostic process. It utilizes a hybrid classification and LLM system to help engineers identify and create "anomalous windows" for faster troubleshooting. This solution is currently enabled for 76 of the company's most active products, which account for an estimated 240,000 cases involving log files annually. By utilizing Analyze SW Support for the past 9 months and across 76 products initially, IBM Support engineers experienced an average time saving of 31 minutes per case (an 11% reduction in time spent). This could translate to a projected annual savings of 124,000 hours (or 10,000 hours per month) and over 600,000 minutes saved each month.2
Finally, Automate SW Support, built on watsonx Orchestrate™ and watsonx.ai, can help streamline repetitive tasks such as case follow-ups, updates, closures and cross-team collaboration. When initially launched across 14 use cases to 100 engineers, these automations delivered an average daily savings of 40 minutes per engineer. Now deployed to 2,800 support engineers worldwide, we project a potential monthly savings of 56,000 hours based on those initial results.
In essence, Ask SW Support, Analyze SW Support, and Automate SW Support work together to empower engineers, significantly improve efficiency, and ultimately, enhance customer satisfaction.
The IBM Software Support organization is a global team of IBM Software product experts with decades of experience working with clients across all industries and varying technologies. They are focused on providing assistance to IBM Clients who encounter technical issues or need technical help with IBM Software products. In addition to technical issue resolution, the IBM Software Support team often informs users proactively about software updates, patches, and bug fixes, and collects customer feedback to help drive enhancements. The IBM Software Support team is an extension of the development labs to help design and implement changes, providing a highly resilient service throughout the software product lifecycle.
¹ Based on a projected 10 min productivity gain when correct answer generated (28% current accuracy).
² Based on estimated 240,000 complex, higher severity cases. It is assumed that diagnosis for these complex cases requires log file analysis. Based on a projected average 31 min time savings for 240,000 cases.
Legal
© Copyright IBM Corporation 2024. IBM, the IBM logo, IBM watsonx and watsonx.ai are trademarks or registered trademarks of IBM Corp., in the U.S. and/or other countries. This document is current as of the initial date of publication and may be changed by IBM at any time. Not all offerings are available in every country in which IBM operates.
Client examples are presented as illustrations of how those clients have used IBM products and the results they may have achieved. Actual performance, cost, savings or other results in other operating environments may vary.