Nexa: Automatically Surfacing Business Impacting Insights in E-commerce Applications
Smart Sun (Conviva), Sayan Sinha (Georgia Tech/Conviva), Haijie Wu (Conviva), Aditya Ganjam (Conviva), Qichu Gong (Conviva), Wei Wang (Conviva), Zhan Yang (Conviva), Bo Lin (Conviva), Vipul Harsh (Conviva), Ningning Hu (Conviva), B. Aditya Prakash (Georgia Tech), Vyas Sekar (CMU), Hui Zhang (Conviva)
Architectural Patterns & Composition
Summary
An agentic framework that automatically discovers business-impacting behavioral patterns across billions of e-commerce user interactions using contrastive stateful trajectories.
Description
Internet-scale e-commerce storefronts serve millions of users (and increasingly user appointed agents). These storefronts are being rearchitected as compound AI systems with agentic workflows for customer interactions, backend processing, and so on. As this AI transformation and agentic economy is underway, product teams need to get actionable insights into business-impacting outcomes. Classical approaches for understanding business impacts such as static funnels or static dashboards cannot deal with the scale, diversity, and contextual interactions that happen over billions of user interactions. As such, we need novel agentic approaches to automatically surface business-impacting insights. We present Nexa, an agentic framework that surfaces business insights automatically. We formalize the target of automated insight discovery in terms of Contrastive Stateful Trajectories (CST): a structural specification over contextual and sequential behavioral patterns whose presence or absence significantly shifts a business KPI across user cohorts. Nexa satisfies three design requirements simultaneously: expressivity through the CST abstraction, scalability through a novel custom analytics backend for CST computations, and explainability by using the CST abstraction and usable presentation layers for analysts to verify the insights. We demonstrate Nexa on representative ecommerce workloads and show that it surfaces high-signal friction contextual patterns spanning user, app, agent, and backend behaviors.
