William Whitty

A 2D game in a dynamic universe exclusively governed by LLM chat models like ChatGPT and Llama.

I led the game's development and implemented innovative game mechanics, enabling the autonomous generation of character sheets, star systems, and planets through advanced chatbots. I also designed an intricate AI-driven character system that facilitated in-game entities to interact with players and remember these interactions for a personalized gaming experience.

Additionally, I developed and integrated procedural generation algorithms to create a vast and diverse cosmos featuring unique star systems and planets influenced by the ChatGPT-controlled environment. I engineered a dynamic narrative experience by utilizing chatbots to generate and control non-playable characters, ensuring evolving story arcs based on player choices. I also established a player-centric approach with persistent consequences, allowing decisions to shape the in-game world, including political landscapes, economies, and social dynamics.

A sophisticated financial charting system using the Godot game engine. I led the design and implementation of this state-of-the-art system, which harnessed the power of GPU shaders to render up to 100,000 candlesticks seamlessly and concurrently, providing an unparalleled visual representation of market data.

I developed a pluggable add-on system for indicators, enhancing the system's adaptability to different analytical needs. I also implemented support for multiple screens and seamlessly integrated various data sources through a custom-built data broker, enabling comprehensive market analysis.

Additionally, I engineered a feature-rich toolbox, incorporating essential functionalities commonly found in professional charting software. I also pioneered an innovative fractal polynomial compression system, a groundbreaking approach yet to be documented in academic papers, showcasing a commitment to cutting-edge technological solutions.

A 3D top-down game I conceptualized and developed, offering players the role of a fox navigating a World War II battlefield set along the Finnish border, immersed in a snowy environment.

I crafted a unique gaming experience where players must skillfully maneuver through the treacherous terrain, avoiding guards and engaging in activities like hunting deer. I also introduced pioneering systems to enhance gameplay. I implemented an innovative AI pathfinding system utilizing depth cameras to dynamically detect walkable areas within the scene, ensuring a realistic and responsive environment.

Additionally, I developed a groundbreaking snow-tracking system, allowing footprints to persist indefinitely. This feature provides a lasting visual record of player and AI movements, enhancing realism and strategic gameplay. Next, I utilized Godot for game development, showcasing proficiency in this platform to create an immersive and visually captivating gaming experience.

Although currently unfinished, the game is a testament to the exploration of novel game development concepts and technologies, offering a glimpse into the potential future of interactive and dynamic virtual worlds.

An innovative IoT application I pioneered for True Digital Park, redefining co-working spaces by optimizing functionality.

I orchestrated the seamless integration of IoT devices, including air conditioning and facial recognition gates, to streamline office meetings. I also implemented an adaptable architecture designed for scalability, explicitly catering to events hosting 3,000 to 5,000 participants. Finally, I engineered a cutting-edge modular microservice structure on the Molecular Node.js microservices framework, allowing the swift generation of client-specific configurations.

I contributed to 85% of the application, ensuring a tailored and streamlined platform aligned with client needs. I positioned the application for resale with a strategic emphasis on modularity, broadening its market reach and appeal across diverse industries.

A robust data processing system, leveraging ZeroMQ for efficient interprocess communication and employing a combination of Python and Nim for implementing microprocesses. I led the development of this intricate system, which incorporated a dynamic orchestration of sequential, parallel, and Saga-based microtasks to optimize data processing workflows.

I designed and implemented a distributed architecture comprising approximately 20 nodes, each specializing in different types of data processing tasks. This architecture allowed for scalability and flexibility in handling diverse data processing requirements. I utilized ZeroMQ to establish a seamless communication pipeline between nodes, ensuring efficient and reliable data transfer across the entire system.

In addition, I implemented a combination of Python and Nim for microprocesses, capitalizing on the strengths of each language to achieve optimal performance and code efficiency. I introduced a sophisticated logging and error notification system that tracked the entire data processing workflow. This system facilitated comprehensive monitoring and provided timely alerts for error resolution.

A game development workshop I initiated for a large coworking and office space, driven by a passion for giving back to the community.

I collaborated with building owners to define project scope and requirements, swiftly assembling a dedicated team from my company. I strategized tasks, workshops, and presentations, tailoring the content to accommodate diverse skill levels. On the day of the event, I drew an impressive turnout of around 60 participants across age groups.

I conducted engaging presentations alongside two colleagues, guiding attendees through implementing changes in a small game prepared for the workshop. I also fueled excitement by offering prizes for accomplishing tasks—witnessing enthusiastic engagement from university students—one of whom innovatively modified the code to gather all stars on the map.

A collection of multiple game development projects I've worked on, which are as follows:

• Engineered two custom engines in Lisp and C++, showcased on github.com/tavurth/gamesavvylisp/.

• Applied Three.js for interactive websites and browser-based 3D games. I demonstrated CLOD terrain implementation on github.com/tavurth/terrain-example/ and link.springer.com/chapter/10.1007/978-3-540-28643-1_78/.

• Contributed to the core engine C++ source code of Godot, which you can see on github.com/godotengine/godot/pulls?q=author:tavurth+is:pr/.

• Participated in multiple game-jam competitions, releasing small prototype projects available on the Itch.io linked to this project entry.

• Engaged in longer-term personal game development projects, including dynamically loading procedural worlds, procedural generation of plants and trees, and shader-based AI and terrain deformation.

A tailored solution I developed for a client aiming to explore the latent space within existing trading algorithms. This ambitious project involved creating a sophisticated framework capable of systematically navigating and evaluating a diverse array of algorithms.

Initiating the project involved developing foundational machine learning models—guiding them through an evolutionary process designed to cultivate and enhance their capabilities.

After an intensive month or two of refinement, I successfully delivered an efficient, cutting-edge progressive machine-learning generator using TensorFlow and AST. This advanced system produced a new and improved trained model schema at an impressive rate of one every 10-15 minutes, seamlessly running on a Raspberry Pi. Subsequently, these models underwent further refinement by the client.

Over several months of continuous operation, the system uncovered insightful nuances within specific markets, demonstrating the developed framework's efficacy in delivering valuable insights for the client.