Tech & Art Lead, Project Manager
VR: Meta Quest 2 & 3
1
~1 Month
Multiplayer Metaverse
A research and development metaverse platform built to evaluate scalable multiplayer VR systems, interactive training workflows, and AI-driven avatars. The project focused on combining real-time networking, immersive training environments, and conversational AI into a unified VR ecosystem.
Overview
Experimental Metaverse Prototype (VR Systems Research):
- Multiplayer networking using Photon for real-time shared VR sessions
- Integration of Meta XR SDK for VR interaction and hardware support
- VR Builder used to design interactive training workflows and scenario-based environments
- Inworld AI integrated for real-time conversational NPCs and AI assistants
- Ready Player Me avatars implemented for customizable and cross-platform user identity
System Goals & Validation:
- Validated modular VR architecture for scalable multi-user environments
- Explored AI-driven interactive training agents in immersive spaces
- Tested persistent avatar-based identity systems across sessions
- Evaluated production-ready performance constraints for VR systems
- Structured as a multi-scene VR hub connecting specialized experiences
Technical Break Down
Contributions
Experimental Metaverse Prototype (VR Systems Research):
- Multiplayer networking using Photon for real-time shared VR sessions
- Integration of Meta XR SDK for VR interaction and hardware integration
- VR Builder used to create interactive training workflows and scenario-based environments
- Inworld AI integrated for real-time conversational NPCs and assistants (Unity integration)
- Ready Player Me avatars for customizable user identity and cross-platform avatar consistency
System Goals & Validation:
- Validated modular VR architecture for scalable multi-user environments
- Explored AI-driven interactive training agents
- Tested persistent avatar-based identity systems across sessions
- Evaluated production-ready VR performance constraints
- Designed as a multi-scene VR hub connecting specialized experiences
Real-Time Multiplayer VR Systems (Photon Networking):
- Implemented real-time multiplayer systems using Photon for shared VR sessions
- Synced player positions, animations, and interactions across multiple users
- Managed network ownership and state authority for VR objects
- Handled session joins, reconnects, and room persistence
VR Training Systems (VR Builder Workflows):
- Used VR Builder to design structured interactive training environments
- Built step-based guided workflows (task sequences inside VR)
- Implemented logic-driven triggers and conditions for user progression
- Created reusable training scenario templates
- Designed interaction checkpoints for evaluation and feedback loops
This enabled non-linear but controlled training simulations suitable for industrial and onboarding use cases.
AI-Driven VR Interaction Systems (Inworld AI Integration):
- Integrated Inworld AI into Unity to enable real-time conversational NPCs
- Built AI-driven avatars with persistent personalities
- Enabled natural language interaction in VR environments
- Implemented contextual responses based on environment state
- Integrated voice-based communication inside immersive scenes
This allowed users to interact with personalized AI trainers and assistants during training scenarios.
Modular Multi-Scene VR Architecture:
- Built a multi-scene VR architecture with a central hub-based navigation system
- Designed a Hub Scene as the main entry point connecting all VR experiences
- Developed a Music Scene (Photon Meta Hub-based) for shared social and audio experiences
- Created a VR Logistics System Scene simulating spatial workflows and object movement
- Built a Production Training Scene using VR Builder for structured training workflows
Each scene was modular and network-aware, enabling seamless transitions between experiences in a shared metaverse environment.
Avatar & Identity System (Ready Player Me Integration):
- Integrated Ready Player Me avatars for consistent identity across sessions
- Implemented real-time avatar loading and customization
- Enabled network-synced avatar appearance in multiplayer sessions
- Integrated animation rig system for VR movement and gestures
- Optimized avatar performance for VR hardware constraints
VR Login & Session Management System:
- Designed and implemented a custom VR login system with user authentication flow
- Built session creation and joining logic for multiplayer VR environments
- Developed avatar selection and initialization pipeline
- Created a clean VR-native UI optimized for controller-based interaction
This system acted as the entry point into the metaverse environment.
VR Performance Optimization & Asset Engineering:
- Optimized imported 3D assets for real-time VR performance
- Reduced polygon counts for performance-critical objects
- Baked lighting (lightmaps) for static environments
- Performed material optimization and shader simplification
- Implemented texture compression and LOD (Level of Detail) systems
- Ensured stable frame rates across all VR scenes
Most assets were externally sourced and then heavily optimized to meet strict VR performance budgets.
Key REsults
Metaverse & VR Systems Engineering Experience:
- Built and evaluated a full multiplayer VR architecture using Photon
- Gained hands-on experience with AI-driven NPCs in real-time 3D environments (Inworld AI)
- Developed modular VR training systems using VR Builder
- Worked with avatar identity systems (Ready Player Me) in multiplayer VR
- Learned advanced VR optimization techniques (baking, LODs, shader reduction)
- Designed scalable multi-scene metaverse architecture
- Understood constraints of real-time networking in immersive environments
Tech Stack
Technical Skills:
- Unity
- Photon Fusion (Multiplayer Networking)
- Meta XR SDK
- VR Builder
- Inworld AI (Unity Integration)
- Ready Player Me
- C#
- XR Interaction Toolkit
- Real-time Multiplayer VR Systems
- 3D Optimization (LOD, Baking, Material/Texture Optimization)