Hi, I'm Tanguy Geraedts
Aspiring Full Stack DeveloperSoftware dev out of KdG with a soft spot for backend work. I spend most of my time building solid systems, experimenting with new tech, and constantly diving into personal coding projects. I just love creating cool software that works well, solves actual problems, and lets me build whatever comes to mind.
Experience
2026March 2026 – June 2026Internship
Java Full Stack Engineer
Roots SoftwareAntwerp, Belgium
Engineered end-to-end features for a global, location-based real-world multiplayer game across a Spring Boot REST API backend and a React/TypeScript production codebase.
Architected an AI-driven game board generation pipeline using LangChain4j, leveraging single- and multi-agent LLM/VLM systems to autonomously generate, validate, and self-correct gameplay content.
Built real-time interactive UI components using Tailwind CSS and WebSockets, backed by Testcontainers integration tests and automated CI/CD deployment pipelines in an Agile Scrum environment.
JavaSpring BootReactTypeScriptTailwindCSSLangChain4jWebSocketsTestcontainersCI/CDScrum
Projects
Tomorrowland Video Proxy Pipeline
A production-grade, distributed video processing pipeline built to process 500 TB of RAW festival footage across 16 stages with zero job loss.
TypeScriptNestJSRabbitMQFFmpegPrismaReactProResH.264
BanditGames (AI Board Game Platform)
A full-scale, AI-powered digital board game platform built on hexagonal microservices, event-driven RabbitMQ messaging, and custom MCTS AI opponents.
JavaSpring BootRabbitMQPythonFastAPIKeycloakPostgreSQLMLflowRAGMCTSReactTypeScriptTailwindCSS
Multi-Agent Debate System
An AI-powered multi-agent debate platform built with FastAPI and local Ollama LLMs, featuring a modern Vite, React, and Tailwind CSS frontend.
PythonFastAPIOllamaViteReactTypeScriptTailwindCSS
RL Moon Landing & Reward Shaping
A Deep Reinforcement Learning study analyzing the impact of custom reward shaping and architecture selection in OpenAI Gymnasium's LunarLander-v3.
PythonStable-Baselines3GymnasiumPyTorch
Fashion-MNIST Vision & Grad-CAM Analysis
A deep learning image classification project comparing a custom CNN with fine-tuned pretrained vision models, featuring Grad-CAM explainability.
PythonPyTorchComputer VisionGrad-CAMJupyter Notebook
Skills
Languages
JavaTypeScriptJavaScriptPythonSQL
Frontend
ReactNext.jsTailwindCSSHTMLCSS
Backend
Spring BootNode.jsREST APIsWebSockets
Data & Infra
PostgreSQLDockerRabbitMQCI/CD
Tools & Practices
GitScrumTestcontainersFigma
Contact
tanguy.geraedts4@gmail.comOpen to freelance work, and full-time opportunities. Feel free to reach out!
Antwerp, Belgium