OOODLE
PoC for a mobile multiplayer roguelite game embedded as a widget in a Flutter-based instant messaging app. Matches are automatically created based on users in the chat room, displaying their avatars as they interact and progress through sponsored biomes. Character leveling is tied to chat activity (e.g., message count, length, media content), while the system awards NFTs for winning fights and unlocking new areas. This proof of concept helped secure investments for the full-featured app.
My contributions
Multiplayer client/serverHot topic 🔥🔥
- Netcode for GameObjects: Syncronization of character location and shared rewards.
- Game Server Hosting: Server configuration and Unity backend development
- Matchmaker: Automatic match creation based on the id of the current chatroom (Shared through Flutter integration)
Level generation
Developed the procedural generation system, following (and contributing to) design specifications. It is achieved defining sockets on tiles for each unique connection and chosing a (weighted) random connection from a pool of available tiles.
Flutter integration
Implemented the Unity side of the Flutter-Unity integration, allowing the game to run as a widget inside the Flutter app. This involved creating a custom UnityPlayerActivity and implementing a bridge for communication between Unity and Flutter. The integration allows for seamless interaction between the two platforms, enabling features such as in-game chat and real-time updates. Documentation: Created detailed documentation for Flutter developers on how to integrate and use the Unity widget within their app.
AI systems
Created a flexible AI system that allows composition and sharing of behaviors. It is based on a hierarchical state machine pattern, with the twist that every submachine is a pluggable and reusable component. The top-level machine is customizable for each character, making it powerful but also easy to use.
Realtime weather
Implemented API calls to geolocalize the user through IP address (if GPS is unavailable) and fetch local weather conditions from OpenWeatherMap. The current weather is then used to load and apply the corresponding camera filters and/or particle systems.