Luke Kulm

Download Resume

About Me

Photo of Luke Kulm

I'm Luke Kulm, a software engineer and researcher from Juneau, Alaska, currently studying Computer Science at Cornell University. My passion lies in machine learning, robotics, and AI safety, where I blend theory with hands-on experience. I strive to write clean code and foster strong collaboration within teams, always pushing myself to learn and innovate. Beyond engineering, I'm deeply interested in the intersection of technology, law, and policy, especially as it relates to AI. Outside of work, you'll find me skiing, fishing, or hiking—staying connected to the outdoors whenever I can.

Skills

Programming Languages & Packages

Python
PyTorch
C/C++
Ocaml
ROS
HTML
CSS
JavaScript

Tools & Technologies

GitHub
Linux
AWS
VSCode
Fusion360
LaTeX

Publications & Projects

MORPHeus project image

MORPHeus: a Multimodal One-armed Robot-assisted Peeling System

Conference: ICRA 2024

Collaborated with other researchers to develop an autonomous food peeling system that integrates a robotic arm with multimodal perception and natural language processing. Leveraged technologies including ROS, Python, PyTorch, Linux, CAD, and GPT-4, and co-authored a research paper on the system.

ROS Python PyTorch Linux CAD
Robot Learning from Human Preferences project image

Robot Learning from Human Preferences

Spring 2024

Implemented a Deep Reinforcement Learning from Human Feedback (RLHF) pipeline to train robotic behaviors using human preference labels. The system learns robotic manipulation tasks without explicit reward functions by leveraging human comparisons between trajectories. Built a modular implementation of the RLHF algorithm with a complete pipeline for preference collection, reward modeling, and policy training using PPO in a simulated environment.

Python PyTorch Reinforcement Learning Robotics
Adversarial Robustness of I-JEPA project image

Adversarial Robustness of I-JEPA

Fall 2024

Evaluated the adversarial robustness of Meta's I-JEPA by testing its ability to embed images under adversarial conditions.

Python PyTorch Computer Vision
Reinforcement Learning Visualization photo

Reinforcement Learning Visualization

Spring 2024

Developed an interactive web-based visualization tool for reinforcement learning algorithms, helping students understand complex RL concepts through dynamic visual representations. The project combines deep learning implementations with a web interface.

Python PyTorch JavaScript HTML CSS
Personal Portfolio Website photo

Personal Portfolio Website

Spring 2024

Designed and developed a personal portfolio website from scratch, showcasing my projects and skills. The site features a modern design with smooth animation and responsive layouts.

HTML CSS JavaScript GitHub Pages

Cornell's First Deep Learning Class for Undergraduates

Fall 2023

Created curriculum and course content for Cornell's inaugural undergraduate course on deep learning, CS 4782. Designed programming assignments and homework on deep reinforcement learning, contributing to hands-on education in cutting-edge AI.

Python PyTorch Deep Learning LaTeX
BrainRobotConnect project image

BrainRobotConnect (TreeHacks 2024)

Spring 2024

Developed a robotic pipeline that utilizes brain signals to control robotic systems, incorporating signal processing and data analysis techniques with Boston Dynamics' Spot robot. This project explored the intersection of neurotechnology and robotics.

Python ROS Signal Processing Robotics
Conway's Game of Ocaml project image

Conway's Game of Ocaml

Spring 2023

Developed a cellular automata simulation using Ocaml, GitHub, and functional programming principles. This project is an implementation of Conway's Game of Life and highlights my skills in functional programming and algorithm design.

OCaml GitHub Functional Programming

Experiences

Education

Cornell University

May 2025

Bachelor of Science in Computer Science

GPA: 3.71

Relevant Coursework: Machine Learning, Analysis of Algorithms, Robot Learning, Computer Vision, Large Language Models, Systems Programming

Work Experience

Yrikka

Summer 2024

Machine Learning Engineer Intern

  • Designed and implemented a scalable evaluation pipeline to assess the robustness and accuracy of multimodal LLMs, focusing on vision-language capabilities in healthcare
  • Leveraged deep learning techniques and LLM agents to benchmark model performance
  • Employed adversarial attack strategies to test the resilience of vision models against perturbations
Python Pytorch AWS

Cornell University

Spring 2024 & Spring 2025

Teaching Assistant - CS 4782: Deep Learning

  • Created and refined programming assignments using PyTorch and data analysis techniques across topics such as Diffusion Models, Vision Transformers, and Reinforcement Learning
  • Assisted students with concepts across deep learning development and theory
Python Pytorch LaTeX

EmPRISE Lab

Spring 2023 - Spring 2024

Undergraduate Researcher

  • Implemented Python-based machine learning techniques to classify video, audio, and haptic data
  • Employed ROS and Linux to orchestrate data acquisition and robot control; rapid prototyping with Arduino and CAD technology to fabricate a vegetable peeler for robot and human use
Python Pytorch ROS