Support teams handle thousands of customer cases, but identifying the specific product issues driving that volume is difficult. Traditional categorisation often lacks the detail needed to help product and engineering teams pinpoint what needs fixing, while the business impact of those fixes can be hard to demonstrate. In this session, the speaker shares how AI-powered clustering and large language models can turn unstructured support data into actionable product insights. He will explore how to uncover recurring issues, identify underlying friction points such as bugs and documentation gaps, and quantify their financial impact to help teams prioritise improvements. Key takeaways: - Use AI to uncover hidden patterns and emerging product issues across large volumes of support cases. - Move beyond broad case categories to identify the specific problems driving customer friction. - Combine support team expertise with AI-generated insights to identify potential solutions. - Quantify the cost of recurring issues and engineering overhead to build a stronger business case for product improvements.