AI-Powered OMS
App

AI OPS needed to turn PDF orders into structured information for an existing business workflow. Orders could contain product names, quantities and specifications across several pages. The system needed to extract those details and match the descriptions to an internal product database.

Reading an order line and identifying its product are separate tasks. The application organizes them into upload, review and export. Staff can examine extracted values, compare suggested matches and make corrections before the order data moves to another system.

Fron Hasani led UI and UX design and TypeScript frontend development, handled API integration, and contributed to AI and backend implementation with collaborators. His responsibilities included how extraction and matching results appear and how review controls connect to the underlying services.

Users upload a PDF and review product information in a table. Suggested database matches include confidence information, side-by-side details and visual status cues. Staff can correct mismatches before downloading the prepared data or passing it through the configured ERP connection. Confidence scores support review; they do not establish that a match is correct.

The delivered application connects PDF intake, extracted order data, product suggestions and an ERP handoff. It gives staff a structured place to inspect and correct order information before export, keeping automated suggestions separate from the approved data used downstream.