Partnering with DXTR to develop an MVP for a real-time VR firefighter training system, focusing on skill deviation detection using headset telemetry.

• Developed an LSTM-based model to predict and detect skill deviations based on the last step of the sequence.
• Designed and architected an LLM-driven coaching layer that interprets pose deviation data to provide immediate, natural-language feedback to trainees.
• Engineered a redundant full-body pose collection pipeline using Meta Quest SDK to ensure cross-platform compatibility and reliable telemetry capture.
• Collaborated on data integration and evaluation of pose deviation models for high-fidelity training scenarios.

Core contributor to DreamForge, an AI-driven game engine that procedurally generates complete 3D game worlds from text prompts.

• Implemented a procedural wall cutout system using CSG boolean subtraction with four interchangeable silhouette strategies (convex hull, radial ray casting, voxel boundary tracing, alpha-shape concave hull), BFS connected-component filtering, and ear-clipping triangulation for mesh extrusion.
• Built CI/CD pipelines using Docker, Azure Container Instances, .NET, GitHub Actions, and Slack webhooks for automated exception analysis, PR creation, and deployment notifications.
• Architected a multi-mode camera system using Cinemachine with adaptive framing optimization, multi-candidate camera selection with occlusion filtering, bone-based NPC framing, 3-ray spherecast occlusion fading, and dynamic room-geometry yaw constraints with velocity damping.
• Designed a polygon-based room shape system with grid-based flood fill validation and geometry utilities (segment intersection, polygon scaling, convex hull).

VR and motion-capture projects using Meta Quest, ARKit, Unity3D, and Unreal Engine.

• Designed experiments for VR illusions in the Cubics project and developed the Deep app for immersive iOS-based telecommunication.
• Taught Java-based object-oriented programming labs for four semesters.

Simulation systems for high-traffic environments such as airports, train stations, and malls using Unity3D.

• Integrated pathfinding algorithms to enable adaptive and efficient crowd navigation.
• Created tools to visualize and modify agent behavior in real time for simulation control and testing.

Led development of Techu, a strategic card game — gameplay, AI, and multiplayer integrations.

• Integrated online multiplayer using Photon and PlayFab, improving engagement and online interaction.
• Developed an AI agent using reinforcement learning principles.

Volunteer full-stack developer for IAESTE (International Association for the Exchange of Students for Technical Experience).

• Developed and maintained web applications for non-profit projects with Vue.js and Django, improving functionality and user experience.

Technical staff for AUT Game Development Events for two consecutive years — organizing technical aspects, content, and tutoring — plus teaching assistant roles across several courses.

• Organized and managed event technology, prepared participant materials, and tutored attendees to promote game development education and community.
• Assisted in teaching Principles of Artificial Intelligence, Advanced Computer Programming with Java, and Fundamentals of Computer Programming with C.

Internship at Sepantab, an IoT startup, during the first summer after the pandemic began. Built an online mobile game for café customers to use while waiting for orders — simple engagement and light rewards while orders are prepared.

Game developer on a small team shipping hyper-casual and adventure games for iOS, Android, and Windows.