Data Scientist • Applied AI Engineer
I build production ML systems, autonomous AI agents, and mathematical optimization models that solve complex business problems. Bridging advanced analytics, scalable cloud infrastructure, and measurable ROI.
McGill University
Desautels Faculty of Management
Master of Management in Analytics (MMA)
2023 – 2024 · Montreal, QC
University of British Columbia
Bachelor of Management, Finance (Honours)
2015 – 2019 · Kelowna, BC · 3.7 GPA, Dean's List
I'm a data and operations guy who gets genuinely excited about automation. My thing is taking the boring, repetitive work off people's plates — the manual scheduling, the payroll calculations, the copy-pasting — so they can focus on the decisions that actually matter.
I use whatever tools get the job done: Python, AI agents, React, SQL, Databricks. And if I don't know how? I'll figure it out. That's probably my biggest strength — I've never met a problem I couldn't eventually solve, even if it meant learning something entirely new to get there.
Outside of work I'm building Lego, dabbling in options trading (most of my money is safely in ETFs, don't worry), and reading when I can — always open to book recommendations. I studied finance and worked in wealth management, so investing genuinely fascinates me.
Oh, and I'm a bit obsessed with the Artemis program. I actually flew out to see the launch in person — they scrubbed it, obviously — but that's me in the photo with the rocket. No regrets.
End-to-end applications solving real analytics, scheduling, and AI challenges.
A high-performance financial and operational analytics SPA. Renders 2 years of simulated transactional data with zero lag, featuring client-side KPI filtering, product category drilldowns, and customer retention segmentation.
Mathematical constraint programming algorithm in Python (OR-Tools CP-SAT). Generates globally optimal weekly shifts across 10 locations and 50+ employees in under 3 minutes, honoring availability, labor laws, and shift fairness.
Autonomous multi-stage LLM data analysis pipeline equipped with a business data manifest to safely query Gold-layer warehouse tables. Synthesizes findings into HTML reports and slide decks with human-in-the-loop validation gateways.
Multi-facility ERP application built on React, Vite, and Cloudflare Workers & D1 SQL. Tracks stock depletion via Bill-of-Materials (BOM) recipe conversion, with an automated PDF parser extracting supplier invoice items directly into the ledger.
Explore our complete collection of high-performance internal web apps and data tools.
Director of Data & Operations · June 2020 – Present
Architected and deployed the complete data platform, operations research engine, and cloud ERP ecosystem across 10 locations in Europe, serving 50+ staff — built and deployed entirely solo.
Developed an optimization algorithm in Python to determine the most optimal weekly shift assignments across 10 locations and 50+ staff, balancing employee availability, labor rules, and store coverage — cutting a 2-week manual planning process to under 3 minutes.
Architected an automated data warehouse on Databricks with nightly GitHub Actions pipelines syncing POS transaction feeds; engineered automated reporting pipelines delivering product health and category trends across 100+ menu items, saving 15+ hours/week and guiding 4 seasonal menu updates/year.
Built a multi-facility inventory web application on Cloudflare Workers and D1 SQL with automated purchase order tracking and Bill-of-Materials (BOM) recipe depletion; end-to-end automation eliminated the need for an inventory coordinator, saving 1 full-time role's worth of labor (~160 hours/month).
Deployed a custom Streamlit payroll app calculating tiered wages and bonuses directly from shift data (saving 30+ manager hours/month); instituted daily POS cash reconciliation with anomaly detection that flagged 3 cash discrepancies for CCTV review, uncovering internal theft.
Data Scientist · March 2026 – Present
Toronto, ON (Remote)
Architected an end-to-end LLM analysis pipeline equipped with an enterprise data manifest to safely query Gold-layer warehouse tables, using custom analytical skills to automatically generate HTML dashboards and slide decks; built step-by-step human-in-the-loop review gateways, accelerating operational ticket turnaround from 3–5 business days to 3–5 hours (~90% cycle time reduction).
Developed an automated dynamic pricing ML system combining modular regression models predicting price-demand elasticity, return probabilities, cancellation rates, and net profit margins; automated 100% of daily repricing decisions across 400,000+ active SKUs on Amazon, Walmart, and other major marketplaces, eliminating manual pricing intervention while safeguarding profit margins.
Business Analyst & Payment Analyst · 2021 – 2026
eCommerce · Furniture
Toronto, ON
Built an AI agent using Databricks Genie enabling 10+ VPs and Senior Directors to query live warehouse data in natural language, slashing ad-hoc insight turnaround from 2–3 days to instant self-serve queries; implemented human-in-the-loop observability using MLflow to track every query and user interaction, systematically auditing failure cases and rolling out continuous schema and prompt improvements to ensure response accuracy.
Designed and led an end-to-end A/B test evaluating CGI vs. studio photography; presented statistical findings to executive leadership, saving $800K+ USD annually in marketing production costs with zero impact on conversion.
Developed an in-house cross-sell attachment rate algorithm analyzing 4,000+ active SKUs, replacing a third-party vendor to save $20K+/month ($240K+ annually) with an automated daily Tableau feed.
Reviewed and labeled 500+ high-risk transactions weekly in Sift Science to train fraud ML models, improving transaction risk classification accuracy and minimizing false-positive order cancellations.
Managed dispute evidence workflows across Stripe, Affirm, and PayPal, achieving a ~70% dispute recovery rate and recouping $50K–$100K+ USD annually.
Open to Data Science, Applied AI, and Machine Learning engineering opportunities.