JSmith

About Me

MSc. Artificial Intelligence student

An early-career developer with an interest in game development and AI research. I am currently strengthening my skillset in Unity and C# through personal projects, academic work, and hands-on experience. I am interested in gameplay systems, technical design and the role of code in creating smooth user experiences and satisfying game feel. I also enjoy exploring 3D art and creative tools alongside development.

Skills

• Microsoft Certified: AI-900 (Azure AI Fundamentals)
• Skills City Unity Developer Bootcamp Outstanding Graduate
• Java
• Python
• SQL
• AWS
• Selenium
• Blender

Education

MSc Artificial Intelligence (Creative Stream) || Queen Mary University, London – September 2024 - Present

• Modules: Machine Learning, Applied Statistics, Deep Learning for Audio and Music, Ethics and Law in Advanced Digital Information Processing, AI in Games

**BSc (Hons) Computer Science (Artificial Intelligence) with Professional Practice

Brunel University, London -July 2023**

Work

AImploy – Machine Learning Engineer – July 2023 – -AWS Services (transcribe), Python, LangChain 85% accuracy in emotional state detection -NLP algorithm development increasing recruitment process by 75% -Processing large datasets for model training for ideal candidate selection -Using AWS for cloud processing user data, backend service integration

Brunel University London - Educational Software and Data Engineer – August 2021 – September 2022 -Support the CEDPs team when managing events -Creating Technical Documentation such as test plans and help guide for new systems -Testing and evaluating systems and proposing new ideas -Website creation using back and front-end programming languages (JavaScript, PHP) -Web Scraping using the Selenium tool to identify weaknesses in the University’s code runner examination system -Develop TCP/IP Client and Server connections in order for systems to communicate over a network -Part of the Java SDK development team for the Raspberry Pi Robot that will assists 1st years in the programming module.

Benetalk – July 2020 – September 2020 – Signal Processing & Machine Learning Engineer -As part of the Hardware Team, I had tasks varying from creating and testing scripts in Python and R to analyse speech samples and creating visual representations of the results. Azure ML tools were using to predict stuttering in samples with an accuracy of 92%. -Testing the Benetalk app that was in Beta development at the time, providing feedback and suggestions via Excel. -Using Powershell to see how an algorithm interpreted the samples it was given. -Presentation about the work I have done in the past 3 months. -Jira, Slack and GitHub used as methods of communication whilst developing

Projects

Hospital Simulation Unity-based hospital simulation game modeling real-time patient flow through departments Focused on creating a streamlined, educational gameplay experience inspired by real-world triage systems. Interactive UI, character AI, resource tracking and allocation

Genre and Decade Classification System Developed and evaluated four CNN models (genre-only, decade-only, multi-output, and feature-feed) to classify music tracks by genre and decade using the FMA-Small dataset. Achieved up to 85% genre accuracy and 70% decade accuracy, with analysis showing limited benefit from combining genre and decade predictions. Conducted case studies with real tracks to assess model generalization, highlighting challenges in temporal classification.

Reddit Sentiment Analysis Overall sentiment of Reddit posts comments using Python libraries, SQL and React.js

QMLAC Chatbot Designed the official chatbot for the university’s Legal Advice Centre, integrating NLP, Node.js, and Python backend.

Final Year Project Using Kinect for Windows in order to navigate a RaspberryPi robot through a maze via gesture recognition. Developed using C#, ASP.NET for Naïve Bayes and Random Forest classification of gesture instance and Java for remote TCP/IP control of the robot.

Brunel University Swiftbot Contributed to the design of the university’s RPI programming module, focusing on software testing, PowerShell, Selenium, and documentation for back and front-end development via PHP, Wordpress, Javascript, HTML.

Move, Shoot & Avoid 2D Game Created using HTML 5, Javascript, uses random enemy generation, sound effects and collision detection