Skip to content

Applied AI

VeritasAI

Turns a noisy headline stream into a composed front page (a lead story, supporting coverage, and an insight layer) rather than a plain list of links.

Illustration. Overlapping newspaper clippings on a desk, with a passage on one sheet and a passage on another each tagged as a separate source and joined by a thin line.

What it does

  • Editorial layout with a lead article, secondary rail, ticker, briefing card, and a deeper coverage grid.
  • Insights section combining a word cloud with Plotly sentiment and keyword visuals.
  • The generated Colab notebook covers the fuller pipeline: live Google News RSS retrieval, headline cleaning and enrichment, sentiment analysis and ranking, topic reasoning, translation support, and FastAPI wiring.
  • The repository is intentionally lean: it holds the frontend template and the Colab notebook artifact rather than the full working tree.

Built with

  • Python 3.10+
  • Plotly
  • FastAPI
  • Google Colab
  • HTML/CSS/JS

← All projects