Profile

Hey I'm Felix!

Education

Bachelor of Science in Computer Science
Cornell University, College of Engineering
August 2023 - Expected May 2027

Relevant Coursework

Applied High-Performance and Parallel ComputingIntroduction to Analysis of AlgorithmsFoundations of AI Reasoning and Decision makingIntroduction to Machine LearningIntroduction to Deep LearningDiscrete StructuresLinear AlgebraEmbedded SystemsDigital Logic and Computer OrganizationObject-Oriented Programming and Data StructuresProbablity and StatisticsFoundation of RoboticsData Structures and Functional Programming

Projects

3-Caml problem
3-Caml problem
For a final project in my Function programming class, my and two of my friends decided to build a 3D physics sandbox for the classic physics 3-body problem. We decided to take it a step further and create certain scenarios and ways you can play around in the sandbox
OcamlFunctional Programming3d graphicsPhysics simulation
Automato
Automato
I participated in a claude builders code hackathon with a couple of my friends where we built Automato, a mobile app that transoforms short form videos (TikToks, Reels, YouTube Shorts) into organized, searchable notes using AI. The user pastes the link and the app handles the rest in the background and generates clean structured notes.
OpenAI APIClaude APIReact NativeFastAPISupabase
BirdSense
BirdSense
I built BirdSense, a bird-feeder camera that spots and identifies birds and their species in real time. It runs YOLO for detection and a PyTorch ViT for classification I trained, (~92% on 200+ species), piping results through AWS (Lambda, S3, MongoDB) to a web app to display the results.
PythonNext.jsMongoDBAWSPyTorchYOLOTypeScriptmachine learningcomputer visionReactweb developmentREST APItransformers
Marble Mayhem
Marble Mayhem
Myfriend and I built a tilt-controlled marble maze game on a FRDM board, we read the onboard accelerometer over I²C and stream state via UART. On the PC side, we created a Python GUI that procedurally generates mazes through recursive backtracking and simulates 2-D physics of the marble.
CEmbedded programmingPythonIMU sensorsI²CUARTReal-time/interrupt-driven firmwaremicrocontrollersserial communication

Experience

Undegraduate Research Assistant
AIRLAB (Artificial Intelligence Robotics Lab) - Cornell Tech
February 2025 - Present
Ithaca, NY

Undergraduate Research Assistant on Multi-Agent Reinforcement Learning for human–robot collaboration in healthcare and beyond. You can see our most recent publication here: https://arxiv.org/abs/2511.14135

MARLbenchmarkingMachine LearningPyTorch
Research and Development Analyst
Cornell DEBUT - Project Team
September 2025 - Present
Ithaca, NY

Worked on a team developing a low cost ADHD diagnostic tool by intergrating computer vision and wearable technology to track behavioral markers. Engineered a pupil-tracking pipeline using a DeepLabCut CNN to extract and derive biomarkers like velocity and vergance. Trained a logistic regression classifier using MLE to map these features to diagnostic probabilities.

Product DevelopmentWearable TechnologyEmbedded SystemsC/C++Computer VisionMachine Learning
Software Engineer Intern
VegaMX
June 2025 - August 2025
New York, NY

Modeled wildfire risk for critical infrastructure building a geospatial pipeline that fuses satellite (MODIS) and climate (GRIDMET) data with infastructure asset locations. Computed pixel-level correlations (e.g., Spearman) between vegetation, temperature, wind, drought, and historical fire perimeters to generate features and prototype a risk-scoring function for buildings and electric-grid assets.

GeospatialxarrayNumPyGeoPandasJupyterfeature engineeringCRS/projection handlingdata visualizationDaskZarrmatplotlib