Conversation Analysis was a broader initiative to evolve Filum’s capabilities beyond basic topic-and-sentiment reporting into a more actionable intelligence layer across both operational and management workflows. I helped shape it around two connected areas: Inner Loop, which turned conversations into structured AI-readable signals, and Outer Loop, which turned those signals into reporting, segmentation, QA scorecards, and customer-level insight.
Inner Loop: A key part of the product was giving teams the flexibility to define their own conversation attributes, rather than forcing them into fixed categories or rule-based tagging. By letting users describe each attribute and its values in natural language, Filum could use AI to classify conversations in a way that better matched each customer’s business context and supported faster operational response.
Outer Loop: Those signals were extended into both conversation-level and customer-level reporting, enriching the experience profile, segmentation, agent performance review, and broader operational and performance reports. This made conversation data more valuable not only for immediate action, but also for identifying patterns, evaluating quality, and understanding customers over time.
Timeline: Q2 2026
Status: Shipped to all customers