PS&Co Data Lab
Jan 2026 – Present
Vancouver, BC
Technical Analyst (Co-op) · Digital Analyst (Co-op)
Full-stack analytics co-op spanning ML pipelines, internal tooling, and client-facing data products — delivering production systems across cloud infrastructure and cross-functional stakeholder environments.
- Sole engineer on a production ML segmentation pipeline: UMAP dimensionality reduction + PAM/k-medoids clustering on mixed categorical survey data (Gower distance), with automated k-selection, silhouette/ARI stability scoring, and holdout validation; layered LLM synthesis for persona naming and strategic narrative, supporting 266K+ responses across 111 projects
- Reduced manual reporting time from 2 hours to under 5 minutes for live USGA championship events by building an end-to-end automated classification pipeline; packaged as a reusable tool adopted by a team of 4 analysts
- Analyzed 28,155 ticket transactions ($6M revenue) in Python, segmenting customers by geography across 642 cities and attributing sales to 31 promotional codes; delivered as a client-facing interactive HTML report with a buyer-density map
- Owned full lifecycle of an internal reporting platform built with Python (FastAPI, pandas, openpyxl), React, and Plotly.js — from architecture through QA, deployment, and analyst onboarding
- Automated 111 satisfaction metrics across 7 survey datasets (n=496) for the 2026 Masters Tournament, applying a 3-tier flagging model to surface critical dissatisfaction areas and synthesizing 300+ open-text responses into structured findings