At AltaML, I spent several months immersed in impactful machine learning projects, tackling real-world challenges at the intersection of data science, software engineering, and public infrastructure. Below, I reflect on some key challenges and insights from my experience.

Document Classification: Challenges and Insights

My first major project involved information classification for a large enterprise client with strict regulatory requirements around document retention. The core challenge was classifying documents with incomplete or partial textual information, often with an extremely limited dataset for training and validation.

Below is a copy of a speed talk that I completed at TRIUMF in summer 2024 as a part of the TRIUMF Science Week. If I get a video of the presentation I will be sure to upload it!

Summary

Particle reconstruction in the ATLAS detector is the practice of associating calorimeter and tracker signals particle and determining which particles caused them, with how much energy, and through what process. One crucial step in reconstruction is cell segmentation where we determine which calorimeter signals belong to which tracks from the inner tracker. Current methods to perform segmentation rely on algorithms that grow with quadratic complexity, posing a challenge for segmentation after the High-Luminosity LHC (HL-LHC) upgrade, which promises a significant increase in the number of collisions per event and, therefore, the complexity of each event. Working under Dr. Max Swiatlowski, I am applying PointNet and other machine learning approaches to cell segmentation in the ATLAS calorimeter.