Retail Demand Forecasting AI Agent
Built at an enterprise consulting firm
Overview
Developed a demand planning AI chatbot built in Microsoft Foundry (Azure AI Foundry), connected to user data stored in Microsoft Fabric. Users interact with a conversational interface to query their retail data, and the agent generates 365-day demand forecasts with interpretive graphs. The chatbot leverages a user data function in Fabric for real-time data access and AI-powered predictions.
The Problem
Retail planners needed to forecast demand across product lines for up to a year ahead, but existing tools required deep technical knowledge and produced static, hard-to-interpret outputs.
My Approach
Built a conversational AI agent in Microsoft Foundry connected to a Fabric user data function. Users ask natural language questions about their retail data. The agent queries Fabric, runs forecasting models, and returns interpretive graphs showing demand predictions for the next 365 days. No technical knowledge required from end users.