Florentin Habermaier

Media Technology and Design Student

About Me

The future is no longer what it used to be. It feels more unpredictable than ever. No wonder: AI is changing how we interact with machines, how we perceive the world, and how we go about our daily lives.

I study Media Technology and Design at the University of Applied Sciences Upper Austria, Campus Hagenberg. As part of my studies, I build web applications, games, interfaces, and interactive 3D experiences. Depending on the project, I contribute to concept development, design, frontend development, and backend development. I particularly enjoy contributing to and leading projects that bring technology, design, and interaction together.

Degree
BSc in Media Technology and Design, expected 2027

Skills

Programming Languages

Frameworks & Technologies

Tools

Project Context Used across Hex'n Catch, Mira, Incremental Tower Defense, Codor, and this portfolio.

Projects

Two low-poly witches flying on broomsticks in the Hex'n Catch poster Hex'n Catch Team Lead / Multiplayer Networking An asymmetric four-player game built in Unity. Three witches travel from tavern to tavern collecting magical potions while a witch hunter races to catch them. Each role has distinct abilities, synchronized across the network with Photon PUN 2. Unity / C# / Photon PUN 2 / Blender / Aseprite Open Project Close Project
In a Nutshell

Hex'n Catch was developed by a six-person team. Three witches use teleportation, broom flight, and transformations to collect magical potions. A witch hunter counters them with abilities for tracking, chasing, and catching. The main technical challenge was connecting these two distinct gameplay loops in a stable four-player match.

My Responsibilities
  • Team Lead
  • Photon PUN 2 Networking
  • Lobby, Role, and Match Synchronization
  • Remote Player Interpolation
Technical Focus

Room management with ready states and random role assignment; buffered interpolation and extrapolation for remote players; and RPCs plus Photon Custom Properties for abilities, scores, and match state.

Project Presentation

The full presentation is available on YouTube.

View on YouTube
Mira upload interface for processing medical documents Mira Frontend Development / UI Design / Backend Collaboration Mira converts medical reports, findings, and scans into traceable JSON data using automatic document alignment, OCR, and Qwen2.5. Medical terms are coded with SNOMED CT and prepared for potential use in FHIR-based systems. Web Application / Python / Document Extraction Open Project Close Project
In a Nutshell

Mira processes unstructured medical documents through a multi-stage pipeline. After automatic alignment and OCR, Qwen2.5 identifies diagnoses, symptoms, procedures, and medications. The extracted information is stored in SQLite, returned as structured JSON, and linked back to its source in the original text. During the semester project, recognized medical terms were additionally coded with SNOMED CT. The output was designed as a standardized data foundation for future integration with FHIR-based systems; FHIR itself was not part of the implementation.

My Responsibilities
  • Frontend Development
  • UI Design
  • Backend Collaboration
Technical Focus

Document alignment, OCR, Qwen2.5, SQLite, SNOMED CT, traceable JSON output, and preparation for FHIR-based downstream processing.

Project Presentation

The presentation shows the full workflow and results of this team project.

View on YouTube
Pixel-art scene with a central tower, a knight, and several enemies Incremental Tower Defense Team Lead / Game Loop Development / UI Design and Implementation A 2D tower-defense survival game in which a central tower must withstand increasingly difficult enemy waves and bosses. Gold earned from defeated enemies is invested in upgrades, while an Arduino controller operates the player character and shop. Unity / C# / Arduino / Unity UI Toolkit Open Project Close Project
In a Nutshell

Incremental Tower Defense combines a 2D survival loop, incremental progression, and a physical controller. Enemies move from multiple spawn points toward a central tower while the player character automatically attacks targets within range. Defeated enemies provide gold for tower health, attack damage, attack speed, life steal, and defensive vines. Timed boss fights unlock new enemy types; defeating the final boss completes the run. An Arduino joystick can replace standard controls for movement and shop navigation, while LEDs indicate game progress.

My Responsibilities
  • Team Lead
  • Game Loop Development
  • UI Design and Implementation
Technical Focus

Unity and C# with scaling spawn rates, boss and upgrade systems, a shop built with Unity UI Toolkit, and serial communication between the game and Arduino hardware.

Codor dashboard showing programming exercises, student progress, and AI feedback Codor Frontend Development / UI Design and Implementation Codor is an AI-powered learning platform for programming assignments. Teachers turn reference solutions into structured exercises; students complete them in an integrated IDE and receive contextual guidance instead of finished answers. The prototype placed third at DevFest Linz 2025. React / TypeScript / Gemini 2.5 Flash-Lite / Monaco Editor / PHP Open Project Close Project
In a Nutshell

Codor was developed by the five-person "Hagenberger Fünfeck" team for the "Next-Gen Education" track. Teachers upload reference solutions, which Gemini 2.5 Flash-Lite converts into an assignment description and structured, editable steps. Students solve each exercise in the integrated code editor and receive contextual guidance from an AI tutor without being shown the final solution. A dashboard gives teachers visibility into submissions and learning progress.

My Responsibilities
  • Frontend Development
  • UI Design and Implementation
  • Task and Step Workflow
  • Learning Progress Dashboard
Technical Focus

React and TypeScript with Monaco Editor; AI-assisted task and step generation via Gemini 2.5 Flash-Lite; and a REST API integration with a PHP/MySQL backend.

Result

Third place at DevFest Linz 2025: The Great AI Hackathon at JKU.

Project Presentation

The full presentation is available on YouTube.

View on YouTube

Contact