Skip to content
This repository has been archived by the owner on Mar 5, 2024. It is now read-only.

Latest commit

 

History

History
209 lines (155 loc) · 9.85 KB

README.md

File metadata and controls

209 lines (155 loc) · 9.85 KB
Raw

klutch

Klutch logo

Table of Contents

What is Klutch?

Klutch is a mobile application that digitizes medical notes, with the aim of saving health professionals time and allowing them to spend it where it counts the most: patient care.

a video demo can be found here: https://drive.google.com/open?id=1ADPvhllSiTYxSDb0GsSC3POdwrNrlbyU

High-Level Architecture

Getting Started

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes. See deployment for notes on how to deploy the project on a live system.

App Prerequisites

  • Android device to run the application

Installation

You can install a live version of the app by downloading the APK onto your android device.

Please ensure that your device allows for downloading apks from unknown sources. If it's not, then follow this guide.

Link to APK here: https://drive.google.com/drive/folders/1GatY3stSDhzZMppQWwpu60rX14f1oXDQ?usp=sharing

Please open and install the above link with your android device.

Development Prerequisites

Before starting please enure you have the latest stable Node release (12+) by running the following

node -v

If not, install nvm and run the following command

nvm use stable

Development Set-up

  1. Clone this repo
  2. Navigate to the app directory: cd klutch/app
  3. run bootstrap script: script/bootstrap
  • if you run into permission denied error, try running the following command:
$ chmod u+x script/bootstrap
  1. Set up backend server
  2. Set up app/frontend

Backend Set-up

Navigate to the app/server/ directory and run the server from command line:

$ npm run start:dev

to view tests, run npm test

Frontend/App Set-up

Now that Backend server is setup, we can run our app!

Open a new terminal window and navigate to the klutch/app/client directory then run the app from the command line:

    ```
    $ ionic serve
    ```

Please note that the CSS will not be accurate on a webview, if you are doing CSS development, try deploying on a real device using DevApp as instructed in the upcoming debugging section.

to view tests, run npm test

Debugging

Both frontend and backend environments have live reloading enabled.

To test native features (such as camera) There are two ways to do this:

  1. Ionic DevApp
    1. Download the Ionic DevApp mobile app from Google Playstore.
    2. Make sure phone and computer are connected to the same network.
    3. Run the following command in the CLI:
    $ ionic serve --devapp
    1. Open the Ionic DevApp on your phone. Your app should show up on the homescreen. Click on it and it should launch your app. If your app still doesn't show up after a while, you can find the IP address and port number on the command line and enter them manually through the menu option.
  2. USB debugging via Visual Studio Code. Follow this guide to live debug the app with VSC.

Deployment

Continous Integration

Using Github actions we have a CI script for the frontend here, and the backend here

Logging and exceptions

Logging is implemented using LogDNA. You can view a screenshot of the dashboard here

Exceptions are tracked using Sentry. You can view a screenshot of the dashboard here

Backend deployment

Done through Heroku, our backend is hosted here https://klucth-app.herokuapp.com/

Heroku git remote: https://git.heroku.com/klucth-app.git

Push your server changes to the heroku git remote to automatically deploy your changes.

App build

Make sure you have android studio installed, for more details, follow "Build for Android" section of this tutorial

You can build the android app by following these commands

$ cd app/server
$ ionic cordova build android

An apk file will be generated in this path: platforms/android/app/build/outputs/apk/debug/app-debug.apk

You can directly install it to your device using ADB

$ adb install platforms/android/app/build/outputs/apk/debug/app-debug.apk

NOTE!

Make sure your path points to the android sdk:

ANDROID_HOME=[path to android sdk]

or run the following command on your current terminal session:

$ export ANDROID_HOME=~/Library/Android/sdk && export PATH=${PATH}:${ANDROID_HOME}/tools && export PATH=${PATH}:${ANDROID_HOME}/platform-tools

Tech Stack

You can refer to the service.yml file to see where some of these services are used

  • NEST.js - A progressive Node.js framework for building efficient, reliable and scalable server-side applications.
  • Ionic - Hybrid mobile application framework
  • Heroku - Platform as a service (PaaS) that enables developers to build, run, and operate applications entirely in the cloud.
  • Firebase - Service that stores and syncs data with a NoSQL cloud database.
  • Google Vision API - Detect texts in images using Optical Character Recognition.
  • IBM Watson - Service that provides NLP extraction functionality from unstructured data.
  • LogDNA - A cloud services-based server-log management and analysis tool.
  • Github Actions - Used to set up a CI action
  • Sentry - Self-hosted and cloud-based error monitoring that helps software teams discover, triage, and prioritize errors in real-time.

Debugging tools

Troubleshooting

  • Please enure you have the latest stable node version installed as per instructed in our Dev Prerequesite section
  • With Heroku's free plan, the site will "shut down" after a certain period of inactivity, so the first request will take some time to go through. You can avoid this by pinging https://klucth-app.herokuapp.com/ directly before testing

Decision Log

  • We decided to integrate our MVP with Cerner EHR because this is the EHR that is used by our main Subject Matter Expert.
  • Third party APIs:
    • We pivoted from using Tesseract to Google Vision because the latter has a higher accuracy for OCR.
    • Instead of using another Google Service, we also pivoted to using IBM Watson for NLP, since the latter has a higher accuracy in this area.
  • Backend:
    • We decided to use NestJS for our backend because it supports TypeScript, which makes it easier to align it with our frontend. Furthermore, it also provides abstraction for common Node.js frameworks (e.g., Express).
    • We also handle all third-party API calls in our backend because Google Vision and IBM Watson are not designed to be called from the client side.
  • For storage and database, we decided to use Firebase because it comes with our Google API subscription. Furthermore, it employs a serverless architecture that allows real-time updates and easy retrieval of data from the client side.
  • For CI, we decided to use github actions and set up a CI workflow. We set 2 different workflows for backend and front end.
  • Finally, we decided to host our backend on Heroku due to its ease of use and to benefit from the loggin add-ons (Sentry, LogDNA)

Resources

We kept track of resources that helped us in this document

Authors

Acknowledgments

  • Julian, our CSC491 instructor, for guiding us through the process.
  • Our 454 teammates Andrea and Jemin for helping us brainstorm the idea