We implemented a configurable GenAI solution that redefined the way complex issues are managed and resolved across channels, products, and business functions. When multimodal data including text, media, and machine data is ingested, our solution harmonizes signals and queries, applies intelligence, and captures a feedback loop for resolution.

We utilized a large language model (LLM) to analyze millions of data points, creating a comprehensive defect ontology for issues and defects. The LLM was supplied with contextual information, including business operations, resolved queries, and existing workflows. Leveraging this context, the LLM developed a multi-layered defect ontology that identified where the most significant and impactful defects were located.

To address the root causes of defects, a root cause ‘driver tree’ is identified for each issue, and mapped to the defect ontology using text classification models. This process facilitated the prioritization and assessment of each defect's value.

Our solution transformed issue management across the organization by establishing a universal hierarchy and categorization that spans channels, products, and business functions.