This thesis presents the development of a conversational tool that enables dynamically generated NPC responses during in-game dialogue. The system integrates multiple components, including large language models, speech recognition, prompt engineering, and character modelling, into a unified pipeline designed for real-time, context-aware, voice-driven conversations. NPC replies are shaped by their personalities, relationships, and a structured memory system that tracks past interactions and emotional states. This approach is particularly suited for VR games, as it eliminates traditional dialogue user interfaces and enhances immersion by allowing players to speak naturally and receive lifelike responses. To evaluate the tool's performance and the quality of player interaction, a prototype game was developed and tested in a VR environment.

The development of open-world video games presents unique challenges in task management, particularly due to the spatial nature of many development tasks such as level design and world-building. Traditional task management tools, while effective for general software projects, lack the ability to represent and manage tasks with explicit spatial context within a 3D game environment. Inspired by Nintendo’s innovative internal solution for The Legend of Zelda: Breath of the Wild, this thesis investigates the integration of spatial task management into mainstream game development workflows.

The primary contribution of this work is the design, implementation, and evaluation of ZEER, a tool that bridges the gap between conventional task management systems and the spatial requirements of game development. ZEER integrates the Unity game engine with Notion, a widely used task management platform, enabling tasks to be created, visualized, and managed directly within the 3D game world. Tasks are spatially anchored in the Unity editor, synchronized with Notion via its API, and enriched with metadata that links their virtual location to their database entry. This approach allows developers to intuitively place, discover, and manage tasks in direct relation to the game environment, improving workflow efficiency and reducing context-switching.

The thesis details the iterative, agile development process of ZEER, including the challenges of supporting flexible database schemas, the rationale for adopting a fixed property set, and the technical solutions for bidirectional synchronization and spatial task placement. Evaluation demonstrates that ZEER streamlines collaboration, enhances spatial awareness of development tasks, and aligns with industry practices while remaining accessible to teams of varying sizes and technical backgrounds.

The results highlight the benefits and limitations of spatial task management in game development, providing a foundation for future enhancements such as richer property support, improved screenshot and UI workflows, and broader engine compatibility. ZEER represents a step toward more intuitive, context-aware task management in the games industry, with potential applicability to other domains involving spatially complex projects.

This paper introduces a system for controlling video game aesthetics through the Space Foundation System, a graph-based formalization of game spaces. This system enables designers to define and shape the aesthetics, atmosphere, and emotional tone of various in-game locations. Additionally, it offers flexibility for developers, allowing customization for domain-specific applications. We propose a unified interface that integrates and manages diverse aesthetic systems, including both visual and auditory elements. In addition to the general framework, we present a concrete implementation for the Unity game engine, followed by a practical demonstrator highlighting the system's real-world applicability.

With technological developments and world aging, Alzheimer's disease (AD) has become a more serious issue, which affects the elderly' cognitive and behavioural health. Like other cognitive impairing diseases, Alzheimer's disease places a heavy burden on the global healthcare system, resulting in a need for innovative and practical care solutions. Understanding the people suffering from Alzheimer's is yet challenging, but it helps find a technology-based approach to care systems. Therefore, this thesis paper not only underlines the challenges faced by Alzheimer's patients daily, but also examines ways to integrate innovative care solutions into their lives.

The focus in this thesis is set on the development and evaluation of a smartwatch-based digital support system, named \textit{"Assist Me!"}. The mobile application aims to guide early to middle-stage patients and their caregivers on how to effectively live with the disease. This application’s primary goal is to give patients a renewed sense of purpose by keeping them positively engaged with the tasks in the application system. Tracking the disease's progression is yet another key aspect that the application covers. Using interactive activities to monitor cognitive functions and activities of daily living will allow the system to provide insight to both caregivers and healthcare professionals. Such an approach goes beyond emotional well-being; while enabling continuous assessment, it aims at extending independence to the patients, improving their QOL, as well as providing support to early intervention strategies when needed. Thus, the solution applies gamification principles, AI-driven capabilities, and cognitive-assisting tools to enhance memory recall, convenience in daily routines, as well as emotional well-being. The components include an AI-generated digital photobook for reminiscence therapy; such dynamic cognitive tests as the EuroTest, PAL, and Clock Construction; a reward-based Click \& Connect system that increases a user's engagement; a medication tracking module powered by Bluetooth beacons; and a weekly MMSE cognitive progression assessment.

The results from an online usability evaluation comprising 38 participants (caregivers and patients) using the System Usability Scale (SUS) showed a strong average score of 78.2, indicating a high level of user satisfaction. The results retrieved from the survey show the system's potential to provide accessible, personalized, and engaging supportive care for Alzheimer's users and indicate other dimensions for improvements, such as multilingual functionality, enhanced help systems, and items of customization of those services.

To sum up, this research reflects wearable technology and digital intervention applications on meaningful quality of life-enhancing use for people with Alzheimer's disease.