How I Built an AI App to Handle Expenses in Sigma

Greg Bonnette, Field CTO at Sigma, describes his experience building an AI-powered expense management app called ExpenSigma within a single day, which addresses the inefficiencies typically associated with traditional enterprise expense reporting tools. The app integrates receipt uploads, automatic data extraction via Snowflake's AI Extract function, and approval workflows, all operating directly on Sigma's platform, thereby ensuring that expense data remains connected with the rest of the business's data. This setup enables real-time analysis without the need for data extraction or separate data engineering processes. Bonnette emphasizes the importance of the app's user experience improvements and highlights how Sigma's features, such as file uploads, Actions framework, and row-level security, facilitated a seamless build process. He also explores the addition of a Sigma Agent, which enhances the app by allowing users to automate tasks and check for policy violations, demonstrating the ease of integrating AI agents into workflows. Looking ahead, Bonnette sees potential in expanding interactive capabilities via platforms like Slack and leveraging vibe-coding support to expedite app development, encouraging organizations to begin by addressing their most cumbersome processes.