resume

Bachelor of Science

The Evergreen State College

Biology

Master of Science

Southern Methodist University

Data Science

Data Science Capstone (Client Project)

Southern Methodist University

• Built a multi-criteria decision analysis framework in R, solo, for a confidential global consumer-packaged goods client, scoring the investment potential of about 22,500 brand extensions across a 77-million-row retail dataset.
• Validated against hand-labeled data at 6 to 18 months; median year-over-year case sales growth separated low-performing from viable brands with up to 85% balanced accuracy; ordinal logistic regression reached 93% adjacent accuracy.
• Built a 4-stage R pipeline ending in a self-contained interactive dashboard that stakeholders view locally; presented to the client and about 40 others, with a paper forthcoming in the SMU Data Science Review.

Research Technician III, Jayaraman Lab

Howard Hughes Medical Institute, Janelia Research Campus

• Build semi-automated Python pipelines that process 100+ GB/day of behavioral video, applying statistical analysis and machine learning, including YOLO object detection and tracking, to characterize behavior.
• Directed design of the institute’s first automated Danionella rig, with closed-loop feeders, 3 cameras, projected stimuli, and a hydrophone, built with engineers; co-developed its control software, which a colleague now uses.
• Benchmarked fish trackers against hand-labeled frames, scoring position error, detection rate, and identity swaps, and reduced swaps with Hungarian and greedy trajectory stitching; adding contrastive learning to a YOLO model, in progress.
• Applied nonparametric tests with multiple-comparison correction; all 9 fish learned a closed-loop task to some extent.
• Documented rig workflows; onboarded 2 lab members and trained 1 researcher to run the rig.

Research Technician II, Keleman Lab

Howard Hughes Medical Institute, Janelia Research Campus

• Second author on a 2022 Nature Communications paper on learning and sleep in Drosophila long-term memory; ran the experiments and analyzed the data for four figure panels covering about 900 flies, using permutation and Wilcoxon tests.
• Took over an unpublished dopamine neuron memory project, independently redoing all experiments, analysis, and figures.

Research Technician I, Project Technical Resources

Howard Hughes Medical Institute, Janelia Research Campus

• Processed 20 to 40 dissection samples per day for a large-scale fly neuron screening project, keeping low-quality samples from reaching the confocal microscope; co-authored the 2025 eLife paper on the split-GAL4 driver line resource.