Inspiration
As online shopping becomes increasingly dynamic, customers expect instant support, personalized recommendations, and easy order management. Traditional e-commerce support systems often fall short in delivering seamless, intelligent interactions. We were inspired to create ShopWhiz to bridge this gap—leveraging Salesforce AI capabilities to build a smart shopping assistant that understands customer needs, guides their journey, and handles post-purchase processes like returns with ease.
What it does
ShopWhiz is an AI-driven shopping assistant built within Salesforce using Agent Studio, Data Cloud, and Experience Cloud. It allows users to: Get product recommendations by name, category, and budget View detailed product specifications including price, rating, and features Place orders with customer and shipping details stored in Data Cloud Initiate return requests by providing only an order ID and return reason View and track order and return records Integrate conversational support through channels like Web and Slack (pending permissions)
How we built it
Salesforce Data Cloud: Set up data models for products, orders, and returns. Agent Studio & Einstein: Created intelligent agents capable of handling product inquiries, placing orders, and initiating returns using structured flows. Flows: Built Apex classes and Flows to manage backend logic for order creation, return management, and product recommendations. Experience Cloud: Designed a frontend site where customers interact with the assistant. Slack Integration: Connected Slack as a messaging channel (configuration in progress). Apex Classes: Used Apex to handle SOQL-based product querying and data processing.
Challenges we ran into
Agent Action UI Configuration: Navigating input/output mappings and ensuring they synced with the flow was intricate. Slack Permissions: Encountered permission issues when trying to connect Salesforce to Slack. Data Model Design: Designing a normalized yet efficient structure for Orders and Returns took careful planning. UI Limitations: Some Experience Cloud UI components were unavailable or difficult to configure for in-app messaging.
Accomplishments that we're proud of
Built a fully working AI agent that handles the entire e-commerce flow: recommend → order → return. Successfully configured Data Cloud to store and relate customer, product, and return records. Developed reusable flows and agent actions that align with Salesforce best practices. Deployed a live Experience Cloud site with real-time agent support.
What we learned
Deep understanding of Salesforce AI Agent Studio and its integration with Flows How to use Data Cloud as a central repository for transactional data How Experience Cloud can be turned into a customer-friendly storefront The complexity and potential of channel integrations like Slack, and the security considerations involved Hands-on experience with Apex and Invocable methods in automation
What's next for ShopWhiz - AI-Driven Smart Shopping Assistant
Enable Slack Integration once permission issues are resolved Add Payment Gateway Integration for live checkout Order Status Tracking via chat and email Feedback & Sentiment Analysis after order completion Multilingual Agent Support for international customers Product Inventory Sync from external systems like ERP or Shopify
Log in or sign up for Devpost to join the conversation.