Readbee

Archived Demo

Readbee was an AI-powered ebook reader built with a full ingestion pipeline, custom frontend, and deployed backend on Kubernetes. Here’s a short demo of the v1 product.

Rethinking How We Read

Readbee was a fast, full-featured reading platform built around 90,000 public domain books. It combined high-performance search, a clean reading experience, and scalable infrastructure—all designed to make great literature more accessible, discoverable, and enjoyable.

See it in action

Screenshots & Key Infrastructure

A look at the key infrastructure and how it works.

GitHub commit history
700+ backend commits in production
Kubernetes dashboard showing running pods
Active Kubernetes pods for Readbee
NFS server
NFS server for gutenberg mirror
SQL database
SQL database for book indexing
GitHub commit history
400+ frontend commits in production
Bucket storage
Bucket storage for book images and content
GitHub Actions workflow run
CI/CD pipeline for backend deployments

Built with love.

Tech Stack

We used a lot of different technologies to build Readbee.

Frontend

  • React
    A JavaScript library for building UIs.
  • JavaScript
    The language of the web.
  • Vite
    A lightning-fast build tool.
  • Tailwind CSS
    A utility-first CSS framework.
  • Redux
    Predictable state management.

Backend

  • Ruby
    A productive, dynamic language.
  • Ruby on Rails
    A full-stack Ruby web framework.
  • Sidekiq
    Background job processing for Ruby.
  • Redis
    In-memory data store & cache.
  • PostgreSQL
    Open-source relational database.

Infrastructure

  • Terraform
    Infrastructure as code.
  • Docker
    Container platform.
  • Google Cloud
    Scalable cloud services.
  • Kubernetes
    Container orchestration.
  • Cloudflare
    Global CDN & security.
  • Github
    Code hosting & collaboration.

Lessons Learned

What we learned from building Readbee

We built Readbee, and learned a lot. Here are some of the key lessons we took away.

Execution Mistake

Built Before We Marketed

We focused on shipping a full product before validating demand. In hindsight, go-to-market should’ve come first.

Market Reality

No Real Traction

Despite a polished experience, we couldn’t compete with entrenched reading platforms—discovery and distribution were missing.

Strategic Insight

Reassessed the Market

Analyzing user behavior and competitive offerings made it clear: the opportunity wasn’t ebooks—it was AI-native interaction.

Focused Direction

Pivot to Character Chat

We’re repositioning Readbee around what users actually want—timeline-aware, memory-driven character chat.

What's next?

The next version of Readbee will be a full-fledged AI-powered ebook reader with a timeline-aware chat interface.

Upload any book
Bring your own EPUB, PDF, or plain-text title and start talking to its characters in seconds.
Timeline-aware chat
Jump to any point in the story and converse with characters exactly as they are in that moment.
Automatic character extraction
We parse the text and build personalities for you—no manual prompt engineering required.
Personalized reading memory
Characters remember what you’ve discussed and reference past conversations as you read.

Let's connect

Whether you're curious about the pivot, have feedback, or just want to talk ideas—I'd love to hear from you.

Email meBook a call