Panelyra

Inspiration

Reading manga and manhwa sounds simple until your library starts growing. You end up jumping between different websites to discover new titles, checking what chapter you were on, remembering what you wanted to read next, and trying to find something that actually matches your mood.

We wanted to build something that felt less like a basic manga database and more like a personal reading workspace.

That idea became Panelyra — a clean, intelligent dashboard designed to bring discovery, tracking, recommendations and reading management into one place.

The interface was inspired by modern AI workspaces, where information and actions are always close by instead of being hidden behind multiple pages.

What it does

Panelyra is an intelligent manga, manhwa and light-novel discovery and library platform.

Users can browse and search titles while filtering by origin, format and genre. Panelyra pulls detailed title information from AniList and turns it into a more personalized experience.

Users can:

Discover manga, manhwa, manhua and light novels
Search and filter a large catalogue
Bookmark titles into a personal library
Track current chapters and reading progress
See how many chapters they still have unread
Create a personal reading queue
Quickly continue reading using saved source links
Receive useful reading notifications
View upcoming and tracked titles from one dashboard
Get recommendations based on their favourite genres
Choose different moods to influence recommendations
Customize the interface, layout and appearance
Keep their library and settings saved locally

Panelyra also includes a built-in recommendation assistant that analyses the user's library and genre preferences to suggest titles that better match their taste rather than simply showing whatever is currently popular.

The goal is for Panelyra to eventually become a single home for someone's entire reading life.

How we built it

Panelyra was built as a custom web application with a strong focus on responsive UI, personalization and real-world usability.

For the catalogue, we integrated the AniList GraphQL API, allowing Panelyra to retrieve manga information dynamically while supporting search, genre, country-of-origin, format, sorting and pagination.

We created our own application state and library system to track things such as:

Bookmarks
Reading status
Current chapter
Reading sources
User preferences
Display settings

User library information and settings are stored locally so that Panelyra can remember the user's setup without requiring a traditional account system for the prototype.

For recommendations, we created a scoring system that takes information such as a user's favourite genres, title scores, popularity and selected mood into account. Instead of returning the same recommendations to everybody, Panelyra builds a lightweight taste profile from the user's own library.

We also spent a significant amount of development time building the interface itself. Panelyra contains multiple interconnected components including the discovery catalogue, reader hub, recommendation system, reading queue, notification system, analytics, settings and assistant.

Challenges we ran into

One of our biggest challenges was state management.

Panelyra has many components that depend on the same information. Saving a manga, changing its reading progress or modifying a setting may need to update several areas of the interface at once.

Keeping those components synchronized without making the application unstable required us to rethink parts of the architecture several times.

Another challenge was working with external catalogue data. Not every manga has the same information available. Some titles may have missing chapter counts, covers, release information or external reading links, so we had to design fallback behaviour rather than assuming every API response would be complete.

The recommendation system was another challenge. Simply recommending highly rated manga would make Panelyra no different from an ordinary ranking website. We experimented with combining genre preferences, popularity, scores and user-selected moods to make recommendations more personal.

UI complexity also became a major challenge as Panelyra grew. Adding more features without turning the dashboard into a cluttered interface required constant changes to layout, navigation and component hierarchy.

And, naturally, we encountered plenty of debugging problems along the way — from duplicate interface element keys and rendering problems to keeping components synchronized as the application became larger.

Accomplishments that we're proud of

We're proud that Panelyra evolved from an interface concept into a functioning product with multiple systems working together.

Some of our favourite accomplishments are:

Building a live catalogue using AniList instead of relying on hard-coded data
Creating persistent personal libraries and preferences
Designing our own recommendation scoring system
Building mood-based discovery
Combining discovery, tracking and reading management in one interface
Creating automatic unread-chapter calculations and reading queues
Supporting manga, manhwa, manhua and light novels
Designing graceful fallbacks when online information is unavailable
Building a highly customizable interface instead of a fixed catalogue website

Most importantly, Panelyra now feels like the foundation of a real product rather than simply a coding demo.

What we learned

Building Panelyra taught us that creating a useful application is very different from simply creating individual features.

A feature may work perfectly on its own but become much more complicated once it has to communicate with the rest of an application.

We learned a lot about:

API integration
GraphQL
Application state management
Data validation
Local persistence
Recommendation algorithms
Dynamic interface rendering
Responsive UI design
Debugging larger applications
Designing around incomplete real-world data

We also learned the importance of building around the user rather than simply adding as many features as possible.

Sometimes removing complexity or making an interaction take one fewer click improves the product more than adding another large feature.

What's next for Panelyra

Panelyra is still only the beginning of what we want to build.

Our next major goal is to make the recommendation system significantly more intelligent. Instead of looking mainly at genres and scores, we want Panelyra to understand deeper characteristics such as themes, story structure, character types, pacing and the user's changing reading behaviour.

We also want to add:

More advanced personalised recommendations
Natural-language discovery, such as "find me a dark fantasy manhwa with an intelligent main character"
Better reading statistics and analytics
Reading streaks and goals
Cross-device accounts and cloud synchronization
Improved release tracking
Smarter notifications
Better source management
Collaborative and social recommendations
Mobile-first improvements
More powerful library organisation
Importing existing libraries from other platforms

Our long-term vision is for Panelyra to become more than a manga tracker.

We want it to become an intelligent reading companion — a platform that understands what you read, remembers where you stopped, helps you discover what to read next and keeps your entire reading world organized in one place.