Matheus Kunzler Maldaner

Interactive tool for visualizing and interacting with logic gate neural networks, enhance transparency in AI decision-making.

Implemented Differentiable Logic Gate Networks on FPGAs for increased inference speeds and reduction in power consumption.

Mobile mental health companion app that pairs users with a supportive dinosaur pet, promoting well-being without guilt.

Web application that allows users to upload images and explore how a CNN classifies images based on specific regions, using a saliency slider for feature visibility.

Data analysis presentation using CourseKata data for the University of Florida's largest annual Data Science competition. Awarded best overall project.

This project involved creating a Discord Bot that uses Sentiment Analysis to detect cheating, sentiment, and toxicity in server communications.

AI-powered educational platform that creates personalized learning pathways based on user's background. It secured finalist standing at the 2023 AI Days Hackathon.

A machine learning project focused on classifying brand logos with a 95% validation accuracy, utilizing tensorflow and EfficientNet on UF's HiPerGator supercomputer.

This project involved leading gene analysis of house mouse in R, using machine learning algorithms to decode neuronal network patterns from RNA-seq datasets.

Flutter based app developed to assist in financial planning, featuring dynamic graphs and sorting algorithm analyses for efficient budget management.

A dungeon crawler python game inspired by professor Jason Harrington's love for mathematics developed with pygame for MAD2502.

A software solution for optimizing room assignments at UF, employing C++ and Qualtrics API to streamline the process for over 200 Resident Assistants.