MeepleHelper · formerly DungeonKeeper

MeepleHelper

Executive summary of the case study.

Raúl Álvarez Hinojosa
2 min read

MeepleHelper, formerly DungeonKeeper, helps people prepare and manage board-game sessions. It combines a mobile app, voice assistant, and PDF rulebook search with source references.

From DungeonKeeper to MeepleHelper

The project began as a voice assistant for board games and evolved into MeepleHelper: a mobile app, backend, and dashboard for preparing and supporting game sessions.

Reduce friction around rules

Looking up a rule can interrupt a game. MeepleHelper processes PDF rulebooks and supports rule search with answers linked to their source.

Support a full game session

The mobile app brings together voice, sessions, timers, players, and scoring. The assistant helps resolve questions without pulling the table out of the game.

Separate system responsibilities

The architecture combines React Native and Expo on mobile, a Node.js/Express/TypeScript API with MongoDB and Redis, and a Next.js dashboard. The backend processes PDFs and uses embeddings to retrieve cited rules.

Improve game intake

Recent commits add rulebook import, confirmation of the BoardGameGeek listing, and rulebooks versioned by edition. The repository is private, so I link no code or operational data.

React NativeExpoNode.jsExpressTypeScriptMongoDBRedisNext.jsRAG / PDF

Mobile interface

MeepleHelper game-selection screen with search, favourites and recent filters, and a list of games.
Capture of the project’s mobile interface, showing titles and thumbnails as they appear in the app.

Choose a game

The screen lets players search for board games and review the list before opening a session.

The repository is private, so I do not link code or operational data. For a conversation about the project, email me@raul-alvarez.es.

FAQ#

How does it help with rules?

It processes PDF rulebooks and retrieves answers with references to source content.

What does the architecture include?

A React Native/Expo app, Node.js/Express API, MongoDB and Redis data, and a Next.js dashboard.

Where is the code?

The repository is private. This case shares a high-level summary.

Raúl Álvarez Hinojosa

Raúl Álvarez Hinojosa

AI Architect · AI Solutions Engineer · Head of AI @ Aircury

Profile, projects, and technical case studies on raul-alvarez.es — Granada, EU remote.

More about the author →
© 2026 Raúl Álvarez Hinojosa. All rights reserved.|Privacy