Explore Microsoft Capstone Project

Final project prototype built on Azure for the Explore Microsoft internship at Microsoft Canada during 2022 summer. Key features include the ability to search through a database of bike parts by uploading pictures of serial numbers or similar bike parts, and a custom chat bot that can answer FAQ questions.

Getting Started

1. Creating resources on Azure

You must first create the following resources on Azure in order to obtain the API keys needed:

• Optical Character Recognition
• Custom Vision
• QnA Maker

More information can be found on the Azure Portal

2. Setup environmental variables

Once you have created the above Azure resources, create a .env file in the root of the project and populate it with the following variables

More information on how to get the keys here:

• OCR
• Custom Vision
• QnA Maker

REACT_APP_OCR_ENDPOINT=
REACT_APP_OCR_SUBSCRIPTION_KEY=

REACT_APP_CV_ENDPOINT=
REACT_APP_CV_PREDICTION_KEY=

REACT_APP_QNA_MAKER_ENDPOINT=
REACT_APP_QNA_MAKER_KEY=

3. Testing locally

docker-compose -f docker-compose-dev.yml up --build

Then navigate to localhost:8080

4. Production using Azure App Services

Deploy using Azure App Services with the following configuration:

• Publish: Docker Container
• Options: Docker Compose
• Image Source: Docker Hub
• Configuration File: docker-compose.yml

Alternatively, deploy using the Azure ARM Template