How Fivetran's product team uses AI agents to manage customer feedback at scale (www.fivetran.com)

Fivetran has unveiled a groundbreaking internal tool that leverages AI agents to streamline the management of customer feedback across its extensive portfolio of over 700 connectors. Previously, product managers struggled with the overwhelming volume of feature requests sourced from various channels like Salesforce, Jira, and support tickets. The new customer feedback intelligence platform integrates data from these sources using Fivetran connectors, BigQuery, and dbt, creating a centralized system that enhances the speed and efficiency of request triaging. Significantly, this platform incorporates a multi-agent AI workflow, which includes assessment, browsing for external information, and synthesis of feasibility reports for feature requests. This approach not only accelerates response times and improves prioritization but also ensures timely communication with customers—reducing the cognitive load on product managers and allowing them to focus on strategic decision-making. The lessons learned from this initiative emphasize the importance of clean data, the necessity of aligning AI solutions with specific workflows, and the effectiveness of tailored models for distinct tasks, ultimately transforming how Fivetran’s product team operates.