PersonaReads

Personalized book recommendations based on your reading persona, tone, and themes — not just algorithms.

Comment

Inspiration

We love books — but most recommendation systems feel generic, algorithmic, and impersonal. Whether it's Goodreads, Amazon, or even AI apps, they rarely tell you why a book is recommended — or what it says about you as a reader. We wanted to create a more meaningful way to explore books: not just by genre or ratings, but by emotional tone, themes, and who you are as a reader. That’s how PersonaReads was born.

What it does

PersonaReads is a Chrome extension that offers book recommendations based on your evolving reading persona — not just surface-level genres or popularity.

When a user scans a book title while browsing, PersonaReads:

1 - Retrieves highly relevant recommendations via QLoo’s AI-powered recommendation API, which analyzes taste, content, and metadata to suggest deeply contextual matches.

2 - Uses Gemini through Genkit to:

  • Explain why each recommendation fits, by comparing themes, tone, characters, and style.
  • Analyze the scanned book (and past reading history) to assign or update a personalized reader persona from our curated list of 15 literary archetypes.
  • Track the user’s evolution as a reader over time with persona progression summaries.

It’s not just “people also liked this” — it’s “this book reflects your fascination with moral ambiguity and mythic storytelling.”

How we built it

1 - Frontend: We built a Chrome extension using JavaScript and TailwindCSS to provide an intuitive, minimalist overlay experience while browsing.

2 - Recommendations: We integrated the QLoo API, which uses a combination of metadata, content-based filtering, and taste analysis to return nuanced book recommendations far beyond traditional algorithms.

3 - LLM Integration: For natural language reasoning, we used Google’s Gemini via Genkit:

  • To generate explanations of how recommended books relate to the scanned one.
  • To match the current book to a reading persona archetype based on its themes, tone, and emotional depth.
  • To generate a short, evolving narrative of the user's reading preferences using previously stored personas.

4 - Persona Engine: We designed a custom framework of 15 distinct personas (e.g., The Visionary, The Escapist, The Analyst) to drive deeper personalization and identity-driven discovery.

What's next for PersonaReads

This is just the beginning. Here's where we want to take it next with more features coming in the future:

  • LLM Chat Companion: Let users ask: “Give me something like Dune but more hopeful.”
  • Visual Persona Dashboard: A timeline of evolving personas and book history.
  • Personality-Based Discovery: Let users explore books by tone, theme, and their persona—not genre.
  • Book Clubs & Social Layer: Connect users with similar personas for discussion and group reading.

PersonaReads isn’t just about what to read next — it’s about understanding yourself through the books you love.

Built With

Share this project:

Updates