Gareth Aguilar

- Implement responsive ball throwing through instant-feedback UI and interpolated ball movement
- Programmed dash mechanic through C++, ensuring dash position was always valid by capsule tracing in direction of movement.
- Collaborated with teammates across the world using Git and C++ to avoid merge conflicts.

Dodge Z Ball is a horde style action game where you knockout AI controlled enemies using your dodge ball. This is the first project I completed using a combination of C++ and Blueprints. This game helped me realize the importance of both C++ and Blueprints and why they coexist. Specifically while using Git for source control and having a team of programmers, using a mix of C++ and Blueprints was almost a necessity. As I soon found out, Git and binary files do not merge well together. This meant that only one person could work on a Blueprint actor at a time without losing data. In addition, Blueprints quickly got messy and took up a lot of screen space compared to traditional code. Using only Blueprints was not an option.

My time on the project was primarily focused on player gameplay and mechanics. I developed responsive dodge balls that can pierce and bounce off of enemies based on its state. In addition, I developed a dash mechanic that felt responsive, smooth, and prevented the player from getting stuck in walls through the use of capsule tracing. Finally, I worked on UI and design to polish off the game. I developed a UI system that works with the player mechanics and gives immediate feedback based on the state of the players dodge ball.

- Developed enemy artificial intelligence through the use of Behavior Trees and the Environment Query System (EQS).
- Designed Navigation Queries to support spatially aware AI that allowed enemies to keep a safe distance from the player.
- Programmed custom decorators and tasks to meet requirements for combat-ready AI that supported ranged and melee attacks.

Before working on The Last Gasp, I took a brief break from game jams to reflect back on the last year. I spent the majority of my time using this new-found experience coding Spooky Project, a personal project focusing on modular code and reusing systems rather than being a complete game itself.

Using this time helped prep myself for one of the most ambitious games I helped create to date. I worked alongside another Blueprint programmer to develop a unique action fighter based on possessing the dead. For this project, I was responsible for creating the artificial intelligence in The Last Gasp. Using Environment Queries and Behavior Tree's, I was able to create advanced AI, inspired by Ghost Of Tsushima, that simulated a wolf pack closing in on prey. Enemies would alert others nearby and slowly circle in on their prey. In addition to working on AI, I assisted with other systems such as melee/ranged hit detection and possessing fallen characters.

- Lead a team of 4 and assigned tasks to each teammate to ensure work was distributed appropriately.
- Implemented smooth and responsive player movement through the use of forces in Unity's physics system.
- Developed points system based on player velocity, airtime, and damage taken.

- Implemented AI system that uses audio and visual senses to track the player through Blueprints.
- Develop balanced AI that allowed the player to hide in dark areas and be fully visible in areas with light.
- Responsible for developing interactables and puzzles throughout the dungeon level.
- Coordinated tasks between team members to ensure smooth workflow and prevent merge conflicts.

Following the completion of Beast Arena, I found myself working on a horror style project. The team consisted of two programmers and one designer. This was the first project that I worked on where I did not have to do programming and designing simultaneously. I was lucky enough to be the sole developer of the monster AI.

Using a combination of Behavior Tree's, Blueprints, and the AI Sense Component, I managed to create an immersive boss inspired by Amnesia. The monster featured vision and auditory sensing to determine the player's location. When it encountered obstacles, it was able to break or push through them and chase the player around the level.

- Implemented reactive boss AI that performs attacks based on the location of the player.
- Programmed round system that gradually increased difficulty and introduced new bosses.
- Developed fast paced combo fighting system through the use of Blueprints.
- Designed and programmed drag-n-drop inventory system by taking advantage of built-in Blueprint functions.

Being my first complete project in Unreal Engine, Beast Arena helped propel me to where I am today. Before beginning the project, I was self-taught Blueprints and had little to no experience using the engine. Being the only programmer pushed me to solve my own issues and gave me the confidence to finish the game no matter how many obstacles I ran into. With that being said, I worked tirelessly. Clocking in about 8-12 hours a day focusing on the project, ranging from modeling characters to engineering systems. Looking back, the scope of the game was way over my head and if I had created a simpler game, I could have produced a higher quality project.

Beast Arena went through another pass nearing the end of the year. I spent roughly a month cleaning up the project and applying knowledge from my previous projects. While the game was not reimagined, going through old code and systems helped prove the extent that I had grown as a programmer. I implemented additional comments, refactored old code using pre existing UE libraries to improve stability, and gave the arena level a fresh look to better represent the vision I had of the game.

- Program modular interaction and inventory systems that can be used in a wide variety of genres.
- Design decoupled UI elements so systems do not directly reference them, reinforcing the idea of modular coding.
- Develop basic light sensing system to determine visibility of player and points in the world.

Shortly before beginning The Last Gasp, I took the time to write down what I had learned and compiled it into a collection of modules. That is what this project is. It takes, in my opinion, the best parts of my games in the past and turns it into modules with documentation and comments. The intention is to reuse the code in future projects to speed up development time.

Since this was not a timed project, I focused heavily on quality over quantity and took my time programming all these systems. Including a light detection system, advanced movement component, interaction and inventory system, and a UI module that supports overlapping and removable parts without causing compile errors.

Before beginning my Unreal adventure, I was a Game Maker programmer. I was proficient with the Game Maker Language (GML) and had produced a variety of projects, like Hot Bots, during my free time. Near the beginning of 2020, I was informed of a project that needed a developer to take the lead as programmer.  I was instantly hooked by the art style and premise of the game and have been developing systems for the game since then.

Shortly after beginning work, I did a refactor of the original code to make it expandable and properly documented. I then introduced the battle system, a part of the game where the player solves WarioWare-like minigames to beat the enemy. Near the end of the year, I expanded the project by importing existing libraries to implement a textbox system inspired by Earthbound. Simple grid-based AI was also introduced midway into the project. It took advantage of built-in functions to navigate around obstacles.

- Designed and programmed interaction system and allowed designers to implement new items seamlessly through the use of scriptable objects.
- Implemented PSX-inspired inventory system using C#.
- Programmed flexible UI system that prevented overlapping UI buttons and elements

Shortly after completing Dodge Z Ball, I set my eyes on a new project. What I never expected was to be programming using a completely unfamiliar engine. While I had fiddled around with the Unity Engine, I only had about a month of experience using C# prior to this project. What I soon found out is that programming is less about memorizing syntax and knowing multiple languages, but rather a way of thinking and problem solving.

My first real dive into the engine had me writing foundational mechanics for the project including inventory and interaction systems. I took advantage of Scriptable Objects, a data-based object, to give designers freedom to create items and objects in the world without writing code. In addition, I experimented with events and delegates to prevent tight coupling between classes such as while coding UI and player systems.

- Implement grid-based state machine AI that destroyed barriers and avoided obstacles
- Programmed AI that would actively strafe-dodge to avoid bullets or projectiles
- Develop proof-of-concept multiplayer system using TCP protocol through a server/client infrastructure.
- Implement data-based weapon system to rapidly implement new guns to the arsenal.
- Introduce experimental port to HTML5 to enable gameplay over the browser. 

Hot Bots is a collection of ideas using the Game Maker Studio engine. I had created many prototypes prior but none had come close to the quality and ambition of this game. It could be that same reason that lead to its eventual downfall. Featuring a variety of weapons, fight against a horde of robot killing machines that will gradually get more difficult the longer you survive. Hot Bots was a dream project I developed during my high school years. While it was never finished, I experimented with many tools I never thought I would learn.

During one of the projects phases, it was possible to host and connect to other players using a server-client TCP connection. Enemies, weapons, bullets, and destructible walls would sync over the network seamlessly. In addition, interpolated movement was implemented to fill in for lost or delayed packets. Enemies also featured dodge mechanics that would actively try to dodge the players projectiles and special abilities to attack the player with.

- Implement balanced artificial intelligence that could chase, search, and alert nearby guards of the players presence
- Design and Program grid-based inventory system. The space an item takes in the players backpack will vary depending on size and orientation
- Implement lockdown state. Enemies will trigger a lockdown whenever the player is detected, closing all doors and requiring the player to reset the system.

    After my short break working on personal projects, I decided to participate in another jam. My goal when joining the game jam was to continue improving my AI skillset. I took advantage of Unreal Engine's Behavior Tree system and the Environment Query System. I had experience using both systems in The Last Gasp. My main challenge was balancing and displaying useful information to the player. Similar to various stealth games, enemies had to be predictable and have their intentions clear. This was solved by having a visual status of the enemy and designing clear action/reaction artificial intelligence. 
    Alongside programming artificial intelligence, I also developed the inventory system for the project. The system uses a simple UDataAsset to store item data, the main problem was displaying and handling objects in the players inventory. By visualizing the inventory as a 2D grid of slots, the problem was broken apart and handled in parts. Internally, each slot was either taken or not. When an item was placed in the inventory, it would mark the slot as taken. To take or move an item, each item would store which slots it took up. With the backend finished, visualizing the inventory system to the player was straightforward.

- Implement grid-based state machine AI that destroyed barriers and avoided obstacles
- Programmed AI that would actively strafe-dodge to avoid bullets or projectiles
- Develop proof-of-concept multiplayer system using TCP protocol through a server/client infrastructure.
- Implement data-based weapon system to rapidly implement new guns to the arsenal.
- Introduce experimental port to HTML5 to enable gameplay over the browser. 

Hot Bots is a collection of ideas using the Game Maker Studio engine. I had created many prototypes prior but none had come close to the quality and ambition of this game. It could be that same reason that lead to its eventual downfall. Featuring a variety of weapons, fight against a horde of robot killing machines that will gradually get more difficult the longer you survive. Hot Bots was a dream project I developed during my high school years. While it was never finished, I experimented with many tools I never thought I would learn.

During one of the projects phases, it was possible to host and connect to other players using a server-client TCP connection. Enemies, weapons, bullets, and destructible walls would sync over the network seamlessly. In addition, interpolated movement was implemented to fill in for lost or delayed packets. Enemies also featured dodge mechanics that would actively try to dodge the players projectiles and special abilities to attack the player with.