Home Resume

Work experience

Ephesoft Semantik Invoice

While working at Ephesoft, I played a pivotal role in architecting and designing the front end of Semantik Invoice, a SaaS application designed to automate invoice processing using OCR and machine learning.

Angular Node.js TypeScript AWS UI design Screenshot of the Semantik Invoice app, on the Review page. It displays a scanned invoice from Kord Industries in the main panel, with extracted data and confidence scores in the right sidebar. The left sidebar lists other invoices. Some fields, like the order date and due date, have warnings for low confidence. The interface includes navigation buttons and options to complete or remove the review.

Overview

Semantik Invoice is a SaaS application developed at Ephesoft, designed to automate invoice processing using OCR and machine learning. Customers upload invoice batches, which are scanned to extract key details like invoice numbers, amounts, vendors, and line items. The processed invoices are then reviewed in the web app, where users can confirm or correct extracted data before exporting them into their accounting workflow.

As a full stack software engineer with a front end focus, I played a pivotal role in architecting the front end, improving performance, and enhancing the design and user experience. My contributions included refactoring the front end architecture, designing a reusable component library, optimizing the front end data layer for performance, and implementing critical UI/UX improvements.

What I did

  • Established front end architecture practice
  • Designed and implemented a reusable component library
  • Optimized front end data layer for improved performance
  • Reduced page load times from 5+ seconds down to under 0.5 seconds
  • Took over design responsibilities and implemented critical UI/UX improvements
  • Built a complex table editor for invoice line items
 Screenshot of the Semantik Invoice app, by Ephesoft. The app displays some metrics about the number of invoices that have been processed, as well as a table with details about new invoices that have been uploaded.