Daily Train is a personalized news platform built around worldview-aware reporting.
Users begin with a Political Preferences Personality Test that identifies how they interpret political and social issues. Based on the selected values, the platform generates a corresponding worldview profile and uses that profile to drive personalized news delivery.
- Entry point: a values-based survey ("Political Preferences Personality Test")
- Core output: worldview profile pages (with distinct editorial framing)
- Personalization loop: subscriber-specific agents process incoming news using the subscriber's worldview profile
- Delivery: recurring personalized updates for subscribed users
This architecture creates a consistent lens between user preferences and the tone/framing of the news they receive.
web/- WordPress site and theme code powering the primary product experiencenews/- Python processing scripts for subscriber pipelines and content generation/deliverymobile/- Mobile app surface and integration layer for subscriber experiencesbloke/- Supporting service/runtime scriptsshared/- Cross-surface shared assets and utilitiesinternal/- Internal supporting code and tooling
- A user completes the political preferences survey.
- The system maps selected values into a worldview profile.
- A profile page is created/reused for that worldview.
- Subscriber workflows in
news/process current news through that worldview. - Personalized outputs are delivered to subscribers.
The main front-end experience runs through the child theme in:
web/wp-content/themes/daily-train
For fresh-install setup details (required pages, configuration, secrets, and custom tables), see:
web/wp-content/themes/daily-train/SETUP.md
This project is functional but requires environment setup (API keys, WordPress data model, and custom tables) before full workflows are available.
Minimum path to get oriented:
- Stand up WordPress and activate the
daily-trainchild theme underweb/wp-content/themes/daily-train. - Configure required pages/slugs, secrets, and custom tables using
web/wp-content/themes/daily-train/SETUP.md. - Run and inspect the content processing scripts in
news/to understand subscriber delivery workflows. - Optionally wire the
mobile/app for end-to-end testing of mobile-linked subscriber flows.
- Core personalization flow is implemented and working.
- Fresh-install compatibility has been improved, but setup still requires manual environment-specific configuration.
- Future work is primarily around robustness, observability, and stronger event understanding quality.
Daily Train explores a practical question:
What if people could follow the same events through a worldview that actually matches how they think?
The current implementation is strong in personalization and content generation flow, with ongoing opportunities to improve robustness and event understanding over time.