Alison Thompson Email and Phone Number

I developed HikeIt!, an app that makes finding new hiking trails less time-consuming. Using machine learning techniques, I uncovered 12 topics that are frequently discussed in trail reviews. I then created an interface that allows users to view text from recent reviews related to their topic of interest. Topics include trail conditions, road conditions, wildlife, camping, views, and more.In this project, I scraped 10k+ free-form hiking trail reviews using selenium and BeautifulSoup (Python) and assembled the data into a SQL database. I processed the text using NLP techniques from the NLTK and Gensim libraries (Python) and implemented a guidedLDA model (Python) to categorize text into 1 of 12 relevant topics with 8x improvement over chance. I deployed the app on AWS with bootstrap and flask (Python).

I developed a method to study genetic landscapes leading to relapse in leukemia. This included engineering a novel microfluidic device and creating a custom analysis pipeline to characterize specimens. I extracted 60k+ image features from each of 500+ samples using R (dplyr, ggplot2, k-means clustering) and ImageJ. I also performed hypothesis testing and generated publication-ready visualizations in R. I authored a manuscript published in PLOS One detailing the method and analyzing the failure modes.

I published a new method for gene expression analysis by engineering a novel microfluidic device. I also constructed an analysis pipeline using maximum likelihood estimation to find RNA molecule concentration at 10x greater sensitivity than standard, regression-based methods. I developed devices and assays to diagnose HPV and HIV in low-resource, developing countries and presented results to a team of interdisciplinary scientists. I also served as a teaching assistant for lab and recitation session and mentored 5 honors undergraduate students through experimental design and execution.