Bloom

Bloom is a remote trauma support service from Chayn, a global charity supporting survivors of abuse across borders. Bloom is our flagship product; a free, web-based support service designed for anyone who has experienced or is currently experiencing domestic or sexual abuse. Through a combination of online video-based courses, anonymous interaction and 1:1 chat, Bloom aims to provide tailored information, guidance, everyday tools, and comforting words to cope with traumatic events. 💖

Get involved

If you would like to help Bloom and receive special access to our organization and volunteer opportunities, please get in touch with us to express your interest in volunteering via this form. We'll get back to you to schedule an onboarding call. Other ways to get involved and support us are donating, making an open-source contribution here on GitHub, and supporting us on social media!

Our social medias:

Website - Chayn

Twitter - @ChaynHQ

Instagram - @chaynhq

Youtube - Chayn Team

LinkedIn - @chayn

Bloom Backend

Currently in active development

Technologies Used:

• NestJS - NodeJs framework for building scalable and reliable server-side applications
• PostgreSQL - Object-relational SQL database system
• TypeORM - Object Relational Mapper library
• Firebase - User authentication
• Storyblok - Headless CMS for pages and courses content
• Simplybook - Appointment booking system used for therapy
• Slack - Slack webhooks to send messages to the team
• Rollbar - Error reporting
• Crisp - User messaging
• Docker - Containers for api and db
• Heroku - Build, deploy and operate staging and production apps
• GitHub Actions - CI pipeline
• Swagger - API documentation

Local development

Prerequisites

• NodeJS v16.x
• Yarn v1.x
• Docker

Install dependencies

yarn

Create .env file

Environment variables must be added to a local .env file. Create this file using .env.example for reference, and adding values for local development.

If this is your first time running the system as a volunteer, please get in touch with the team to get access required variables.

Run locally (with Docker) - RECOMMENDED

The project is containerized and can be run solely in docker - both the PostgreSQL database and NestJS app. To run the backend locally, make sure your system has Docker installed - you may need Docker Desktop if using a Mac.

First make sure the docker app is running (just open the app). Then run

docker-compose up

You should see this in the shell output:

Listening on localhost:35001, CTRL+C to stop

Note: you can use an application like Postman to test the apis locally

Run the app

yarn start:dev

You should see this in the shell output:

Listening on localhost:35001, CTRL+C to stop

Database migrations

A migration in TypeORM is a single file with SQL queries to update a database schema as updates/additions are made. Read more about migrations here.

Migrations are automatically run when the app is built docker (locally) or Heroku for staging and production apps.

You'll need to generate and run a migration each time you add or update a database field or table.

To generate a new migration

yarn migration:generate

To run (apply) migrations

yarn migration:run

To revert a migration

yarn migration:revert

New environment variables must be added to Heroku before release.

Run unit tests

To run all unit tests

yarn test:unit

To have your unit tests running in the background as you change code:

yarn test:unit:watch

Formatting and linting

To run linting

yarn lint

To lint and fix

yarn lint:fix

Formatting and linting is provided by ESLint and Prettier (see the relevant configs for details).

Workspace settings for VSCode are included for consistent linting and formatting.

Seed Local Database

If you're a volunteer, the way to add data to your local database is to seed it using a backup file. Please get in touch with the team to get access to the backup file.

If you're staff and have access to Heroku, you also have the option to seed the database via the following script. Before you start, make sure:

1. bloom-local-db container is running in Docker
2. you are logged into the Heroku via your terminal. Read more about the Heroku Cli here
3. Replace <HEROKU_APP_NAME> with the correct Heroku app name in the seed-local-db.sh file
4. Run chmod +x ./seed-local-db.sh in your terminal to make the file executable

After the above has been confirmed, run

bash seed-local-db.sh

Git flow and deployment

Create new branches from the develop base branch. There is no need to run the build command before pushing changes to GitHub, simply push and create a pull request for the new branch. GitHub Actions will run build and linting tasks automatically. Rebase and merge feature/bug branches into develop.

This will trigger an automatic deployment to the staging app by Heroku.

When changes have been tested in staging, merge staging into main. This will trigger an automatic deployment to the production app by Heroku.

Swagger

Swagger automatically reflects all of the endpoints in the app, showing their urls and example request and response objects.

To access Swagger simply run the project and visit http://localhost:35001/swagger

Database models

User Stores basic profile data for a user and relationships.

Partner Stores basic profile data for a partner and relationships.

Partner Admin Stores relationship between a partner and a user, and the partner access records created by the partner admin.

Partner Access Stores the features assigned for a user by a partner. When a partner access record is created by a partner admin user, it is initially unassigned but the unique code generated will be shared with the user. When the user registers using this code, the partner access will be assigned/related to the user, granting them access to extra features. See assignPartnerAccess. Currently there are "tiers" of access determined by the frontend, however if future partners want different features for their users, settings for their features/tiers should be added to the Partner model.

Therapy Session Stores data related to therapy sessions booked via Simplybook, copying the session time and details about the therapist etc. The duplication of data here (vs leaving it in Simplybook) was made to allow therapy reporting to be included in Data Studio. This model also allows for displaying therapy session data on the frontend. Populated by Simplybook -> Zapier webhooks.

Course Stores data related to courses in Storyblok, copying the story id's. The Course records allow us to relate users to courses (via CourseUser) and a Course to Session records, which is required e.g. to check if a user completed a course. The slug and name etc are also stored for convenience, e.g. to be used in reporting. Populated and updated by Storyblok webhooks when course stories are published.

Session Stores data related to courses in Storyblok, copying the story id's. The Session records allow us to relate users to sessions (via SessionUser). The slug and name etc are also stored for convenience, e.g. to be used in reporting. Populated and updated by Storyblok webhooks when course stories are published.

CourseUser Stores relationship between a User and Course records, once a user has started a course. A users progress (completed) for the course is updated (true) when all related SessionUser records are completed for the related Course.

SessionUser Stores relationship between a User and Session records, once a user has started a session. A users session progress (completed) for the session is updated (true) when the /complete endpoint is called.

Key concepts

User types

There are several user types with different features enabled.

Public user - joins Bloom without a partner, with access to self guided and live courses.

Partner user - joins Bloom via a partner (and access code) with access to extra features and courses enabled by the partner(s).

Partner admin user - a partner team member who uses the app to complete Bloom admin tasks such as creating new partner access codes.

Authentication

User authentication is handled by Firebase Auth. Bearer tokens are sent in api request headers and verified in auth.service.ts. The user record is fetched using the retrieved user email of the token.

Crisp profiles

Crisp is the messaging platform used to message the Chayn team in relation to bloom course content or other queries and support. For public users, this 1-1 chat feature is available on live courses only. For partner users, This 1-1 chat feature is available to users with a PartnerAccess that has 1-1 chat enabled.

Users who have access to 1-1 chat also have a profile on Crisp that reflects data from our database regarding their partners, access and course progress. See crisp-api.ts

Reporting

Google Data Studio is an online tool for converting data into customizable informative reports and dashboards.

The reports are generated by writing custom sql queries that return actionable data to Data Studio. Filters are applied in Data Studio allowing data to be segregated into different time periods based on the data createdAt date

License

Bloom and all of Chayn's projects are open source. The core tech stack included here is open source however some external integrations used in the project require subscriptions.