Andrew Boessen



About Me

I'm a Computer Science student at Boston College, with a minor in Finance. I'm deeply interested in software engineering, machine learning, and building innovative projects that solve real-world problems. I am originally from Kansas City and in my free time I enjoy reading, listening to music, and going to the gym. Currently working on a CUDA implementation of Native Sparse Attention and building LLMs with LoRA and RoPE.

Education

Boston College

2022 - 2026

Bachelor of Science: Computer Science

Minor: Finance

Experience

Device Interaction (Intern)

2024

  • Built sleep classification neural network for delivering personalized messages to Garmin Connect users
  • Implemented active learning to train a multi-task classification model
  • Utilized Grafana and Prometheus to monitor performance and drift of model in production
  • Engineered LLM prompts for generating and evaluating personalized messages

Core Platform Technology (Intern)

2023 & 2022

  • Developed tool in Python to monitor usage of ARM Developer licenses across Garmin departments
  • Converted legacy build tool, txt2c utilized by numerous Garmin products, from C to Python
  • Developed grammar for custom file type using ANTLR

Projects

NeuralGameEngine

Neural Game Engine

Neural network approach for modeling interactive game environments using Vector Quantized Variational Auto-encoder (VQ-VAE) and Spatio-Temporal Transformers. Trained on Atari Skiing gameplay data. A detailed technical report is available in the linked repo.

PerfectRep

Perfect Rep

PerfectRep is a 3D pose estimation model tailored specifically for powerlifting analysis. It allows for precise tracking and analysis of lifter's movements to ensure perfect form and technique.

CUDAGrad

CUDA Grad

A tiny entirely unoptimized autograd engine with gpu acceleration built with CUDA and C. A small neural network library is included to trained multi-layer perceptron models with gradient descent by backpropagation.

Native Sparse Attention

Native Sparse Attention

Efficient CUDA implementation of "Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention". This uses warp primitives and Tensor Cores for optimal performance.

EagleEval

Eagle Eval

EagleEval is a platform designed to empower Boston College students with valuable insights into teacher and class ratings. Used by 5,000+ students. Explore at eagleeval.com.

MergeSort

Bitonic Merge Sort

In place sorting of arrays using bitonic merge sort. Built using CUDA to optimize performance for GPU execution. This takes advantage of CUDA's warp shuffle operation to sort numbers in place and in register memory.

ReClip

Re-Clip

Re-Clip is a mobile app that leverages LangChang, MoviePy, Stable Diffusion, and Swift to summarize and generate short-form video content about research papers and GitHub repos through the use of an automated AI pipeline. Built at HackHarvard 2023.

My Bookshelf