Digest Engine logoDigest Engine
DocumentationProjects and Content

Projects and Content

What a Project Is

In this platform, a Project equates to a specific newsletter brand or topic. Everything—Settings, Articles, Team Members, and specifically the AI’s “understanding” of your taste—lives strictly inside boundaries of a single Project.

Content List & Filters

The Content Dashboard displays every article discovered from your plugins.

  • Use the Category dropdown to filter items down to just Tutorials or just News.
  • Use the Relevance slider to hide junk items.

Relevance Scores

The system assigns a score to everything it ingests, representing “How likely is this article a good fit for this project?”

  • Low score (0.0 - 0.4): Probably off-topic.
  • High score (0.8+): Perfect fit.
  • Mid score: Ambiguous. The AI read it and made a guess.

Read more about the math in Algorithms.

Opening a Content Item

Clicking an item reveals the original abstract, AI-generated summary, source metadata, and extracted entities (people/companies mentioned inside it).

Marking Content as Relevant/Not-Relevant

Under each item, you will see a Thumbs Up / Thumbs Down mechanism. This is explicit feedback. Click Thumbs Up only on articles that perfectly exemplify your Newsletter.

How Feedback Affects Future Ranking

When you thumbs-up an article, its underlying data mathematically shifts the Project’s “Topic Centroid.” This means future incoming articles matching that vibe will score higher. See Feedback and Tuning.