Papiro Document Intelligence

• Papiro Document Intelligence
  • Overview
  • Features
  • Tools & Services Used
  • Getting Started
    • Prerequisites
    • Quickstart
    • Text Generation

Overview

As part of Microsoft /BUILD 2022 the AI Platform Group showed a conference intelligence application that used a number of intelligence APIs to showcase what might be done to support conferences with AI. This is the project that was shown!

Microsoft Build free registration is now open and the sessions are listed.

Sign up for the AI session on showcasing this demo: https://aka.ms/MSBuild2022

Features

This sample conference app provides the following features:

• Quick user check-in using ID/Vaccination Card
• FAQ Multilingual Document Translation
• TLDR Policy Document Summarization
• Audio Session Feedback to Sentiment

Tools & Services Used

• Next.js CLI
• Azure Form Recognizer
  • New ID & Vaccination card pretrained models
  • Beta JavaScript SDK
• Azure Speech
  • Speech to Text
  • JavaScript SDK
• Azure Translator
  • Document Translation
  • REST API SDK
• Azure Language
  • Sentiment Analysis
  • Text Analytics JavaScript Client Library
• Text Summarization Preview
• Azure ML CLI
• Bicep is a language for declaratively deploying Azure resources. You can use Bicep instead of JSON for developing your Azure Resource Manager templates (ARM templates).

Getting Started

Prerequisites

A working operating system with Node, the Azure Command-Line Interface(az CLI), and PowerShell or Visual Studio Code and it's Powershell extention

Quickstart

There are a few steps for setting this up:

1. Clone this repo

2. Install yarn globally

   npm i -g yarn

3. Azure Infrastructure setup:

   ./provision.ps1 -name <YOUR_APP_NAME> -location westus
   

4. Azure Web App publish:

   Run the following command to build the web app package and publish it to Azure. It will prompt for login to Azure.

   ./publish.ps1 -name <YOUR_APP_NAME>
   

5. For Web App setup, please refer to /src/README

Text Generation

The text generation API is created using another sample shown at Microsoft /BUILD 2022. If you're interested in setting this up, head on over to the Managed AzureML Inference instructions.

FAQ

Getting Responsible AI error when setup the Azure infrastructure

Users who haven't created Face, Computer Vision, and Language service Azure resources will get the following error when running provision.ps1. Please go to the Azure Portal and create a Face, Computer Vision, or Language service to review and acknowledge the terms and conditions. See more details here.

This subscription cannot create TextAnalytics until you agree to Responsible AI terms for this resource. You can agree to Responsible AI terms by creating a resource through the Azure Portal then trying again. For more detail go to https://go.microsoft.com/fwlink/?linkid=2164190