AI turns untapped data into tech support insights
4 best practices for deciphering complex data
By Militza Bishop | 2 minute read | February 20, 2020
As data continues to grow at staggering rates, enterprises are grappling with how to manage it. They’re sitting on virtual goldmines of tech support insights, yet how to deploy artificial intelligence and analytics effectively can be a struggle. In fact, over 70% of leaders surveyed in a NewVantage Partners study say adoption of AI is a challenge.
IBM AI and analytics leader Milena Arsova said computational power and analytics alone aren’t enough to improve operations. “You need to know where to start in the context of the problem and the business goal, otherwise it’s like buying a sports car in separate parts and having to assemble it yourself,” she said.
Becoming a data-centric organization requires a shift in mindset. Lay the foundation with combined data science and technology expertise and AI methods for data capture and analysis. Some industries such as banking are ahead of the curve and use AI for personalized support or automating repetitive tasks.
Here are four ways to apply AI and analytics to tech support to pinpoint insights, increase productivity and assess the cost impact.
1. Improve uptime and availability for service instances
Stay ahead of issues that can cause disruptions by using predictive maintenance and analytics to assess downtime patterns, predict failures and refine device management processes. Address routine issues that plague your infrastructure, so more time is spent on critical tech support operations.
2. Sharpen root cause analysis
AI helps you find patterns that may not be obvious in volumes of complex data. For example, you might find that 40% of technical support’s time is spent resolving a specific problem. You can look at cause drivers and support models to assess if introducing a different support model such as remote support or a customer-facing smart chatbot will provide quicker resolution at a lower cost.
3. Tackle infrastructure inefficiencies
With AI, you can determine if a technical support device has a higher failure rate and make device refresh recommendations. This can have a direct, positive impact on the overall cost of support and risk exposure to potential failures.
4. Monitor inventory and supply chains
In terms of inventory management, use automation and virtualization to monitor infrastructure health and locate installed bases with expiring warranties or support contracts.
“Enterprise installed bases can range from a few hundred to over 90,000 devices in the case of large international banks, so the potential cost savings of reducing the risk of failure is significant,” said Arsova.
Similar efficiencies can be applied to supply-chain cycles. Analyze tech support devices to identify where there’s a high demand for specific parts. Then tie those devices to parts-distribution centers so you can supply the appropriate stock to meet demand.
Looking to the future
A recent IDC report estimates 41.6 billion IoT devices will generate 79.4 zettabytes of data in 20251. Given soaring data rates, understanding data structure and increasing analytics efficiency are quickly becoming necessities for smarter technical support operations. Digital platforms record every transaction and experience, and data insights can reveal opportunities and competitive advantages.
“AI and analytics need to be continuous, fast and flexible to affect business outcomes,” said Arsova.
1IDC, Worldwide Global DataSphere IoT Device and Data Forecast, 2019–2023, doc #US45066919, May 2019