Images

ChatApp

Technologies and Software

Technologies

Technologies used in PocketMedi is as follows:

FrontEnd:

• Flutter (Dart)
  • Mobile application

BackEnd:

• PRAW

• Twitter Dataset

  • We used these 2 forms of data scrapping to make a dataset full of PTSD and non PTSD data. If we continue this project further, our goal would be to gain access to higher quality data-sets (i.e. actual conversations between PTSD patients and phsyciatrists).

• HuggingFace Transformers Models and Tokenizers

• Tensorflow for machine learning (python)

  • These are the main 2 components (along with the data) to help train the ML model used for this project.

• Django for REST api (python)

  • Converting a HuggingFace Transformer trained with twitter data to be able to output json data for use with lex bot to analyse and give data about patient’s PTSD level. This will be used to diagnose PTSD via the dashboard.

• Amazon Lex

  • Performs autonomous text messaging with the user and will later store the data in an S3 bucket to be later analyzed by Django REST api

Hosting:

• AWS EC2
• AWS S3
• Vultr
  • This where the Django REST api is hosted on
• Vercel
• Firebase

Software

The software used to build and test PocketMedi would be:

Text Editors

• Vscode (live share for collabaraion)
• Neovim (for quick edits)

API Testing

• Insomnia
• Hoppscotch

License

SPDX Identifier: AGPL-3.0-or-later

GitHub

View Github