function track_msdynmkt_websiteinteractions_114741865() { window["msdynmkt"].setUser({ authId: ""}); // ID, e-mail or phone number - see instructions window["msdynmkt"].trackEvent({ name: "msdynmkt_websiteinteractions_114741865", //Trigger title: Website Interactions ingestionKey : "d12470e34c7a4283ac8769a2797687ca-fab4c50f-f357-4516-a3f8-b908fdfbc681-7154", version: "1.0.0" ,     properties: { "dateandtime" : "", "url" : "", "bindingid" : "" } }); } global eia # ID, e-mail or phone number - see instructions user = User(auth_id='') eia.set_user(user) #Trigger title: Website Visit sampleEvent = Event('msdynmkt_websitevisit_083156767') sampleEvent.set_property("bindingid", "") eia.track_event(sampleEvent) AI applied to financial documents using Microsoft Azure
top of page

AI applied to financial documents using Microsoft Azure

📅17th march 2022 - from 10 AM to 11 AM CEST


Financial documents such as invoices, contracts, and receipts contain key information that is of considerable value to the organization.

However, more than 80% of this information is unstructured and it can be a very long process for a human to analyze these millions of documents. Nowadays, thanks to new artificial intelligence technologies we are able to automate the processing of these documents to extract this value and generate insights.

In this webinar, we will showcase a real project that we delivered for which we applied Azure AI to automatically detect errors on thousands of invoices that were not meant to be paid by the company. We will then discover how to:

  • Extract information from invoices

  • Match each item on the invoice versus the ones that are not covered by using NLP

  • Creating a knowledge mining project to consume this information

  • Apply MLOps techniques to industrialize the models


MDW is a Microsoft Partner of the Year in Analytics & AI. We have developed projects applying artificial intelligence to unstructured documents, and we want to share our experience with you during this interactive live event.


bottom of page