Recruitment Dashboard

KPI and applicant dashboard for volunteer recruitment. Companion piece to the volunteer recruitment beta service: volunteer.samaritans.org.

Although the public-facing service is used nationally, this dashboard only concerns itself with applications to branches in a pilot area.

Key features include:

• KPIs for the entire pilot, over the last thirty days, which can be broken down by branch
• A list of recent applicants and upcoming interviews, which can be broken down by branch
• A detailed profile view of each applicant, showing their progress though the journey
• Export data in CSV format for further analysis
• Reschedule applicants' interviews on their behalf

Environment variables

The app needs the following environment variables set to run:

• Standard AWS credentials of AWS_REGION, AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY
• DYNAMODB_TABLE, the AWS DynamoDB table to scan
• DATABASE_URL, a postgres connection string
• IPSTACK_API_KEY, an API key for the IPstack geolocation service
• FRONTEND_HOST, the location of the Samaritans volunteer front-end app on the web
• HOST, the location of this dashboard on the web
• EVENTBRITE_TOKEN, an Oauth API Token from Eventbrite, to calculate some KPIs
• EVENTBRITE_ORGANISER, an Eventbrite organiser ID number, to identify the right account to query
• SESSION_SECRET, to hash session tokens
• MANDRILL_API_KEY, for the Mailchimp Mandrill service, needed to send emails
• GOOGLE_ANALYTICS_ID, the tracking ID for the Google Analytics property

Running locally

To develop locally, you need node and npm installed, and to make sure all environment variables are set. Then:

1. Clone this repository and cd into it.
2. Run npm install and npm run dev.
3. Run npm run migrate to prepare the database tables.

That should start a dev server and gulp, which builds and watches front-end SASS and JS.

Deployment

This app is hosted on Heroku. Changes to the master branch will be reflected automatically on Heroku.

Database import job

To list recent applicants, this app uses a postgres database separate from the live DynamoDB database that applications are written to as they come in.

This is partly because of limitations querying DynamoDB. This doesn't hurt data integrity since that data isn't editable from this dashboard.

The postgres database is synced daily at 4am using a Heroku scheduler job, which runs the following tasks sequentially:

• scan the source DynamoDB table and hold the results in memory
• use those results to seed a postgres database, according to a defined schema

Use npm run migrate to make sure the schema is up to date, and then npm run seed to run the whole job manually.