jengroff / participant-screener

Automated Applicant Screener

Mobile messaging assistant which interacts with, and captures data from, research study applicants.

---

Application Flow

• App responds to inbound SMS or WhatsApp messages from users;
• During a fully automated session, the app:
  • Asks an applicant 3 - 8 screening questions, to help determine study fit;
  • Captures all the prospect's responses, as well as their phone number, email address, and name;
  • Stores it all in a PostgreSQL database;
  • Presents it all in a Web UI for non-technical staff to view, edit, and track.

Moving parts

• Python, YAML, shell scripts
• Flask, SQLAlchemy, PostgreSQL
• Docker, GitLab CI/CD, Heroku
• Ability to manage the Flask app and the database through the CLI
• Pytest fixtures and tests
• Black code formatting

---

Example flow in Twilio Studio

---

Run the app ->

Docker (recommended):

Use docker-compose to build the image and run the container:

docker-compose up -d --build

Create the database schema:

docker-compose exec api python manage.py recreate_db'

Seed the database:

docker-compose exec api python manage.py seed_db'