Open to AI/ML Engineering & Data Science roles

Building intelligent systems
that think, learn, and deliver

AI/ML Engineer & Data Scientist — turning complex problems into verifiable solutions

From supply chain optimization to real-time arbitrage engines, I build systems where outcomes are measurable, decisions are data-driven, and AI isn't just a tool—it's the architecture.

The Journey So Far

Not just code—curiosity, failure, and relentless building.

I started with a simple belief: data should drive decisions, not guesswork. That belief took me from BYU-Idaho's data science program to United Airlines' cybersecurity team, where I reduced pipeline runtimes by 80% and scaled systems 10x.

At WPA Intelligence, I built ML models predicting voter behavior with 90%+ accuracy across 25 campaigns. At Tree Top, I'm designing linear programming models to optimize fruit supply chains—saving real dollars in transportation and warehousing.

But my real passion? Agentic systems. I'm not a "vibe coder." I architect AI that reasons—like Compa, my real-time arbitrage engine that combines LLM reasoning with structured domain knowledge to find profitable flips before anyone else.

25+ projects. Some failed. All taught me something. The ones that worked? They verify their outcomes in dollars, votes, or efficiency gains.

80%
Pipeline Runtime Reduction
United Airlines
90%+
Voter Prediction Accuracy
WPA Intelligence
97%
Report Generation Speedup
WPA Intelligence
25+
Projects Built
2022–2026

Where I've Delivered

Impact measured in numbers, not buzzwords.

Jan 2025 — Present
Supply Chain Data Analyst
Tree Top Inc., Selah, WA
  • Documented full FPGS Excel-based data workflows, identifying pipeline optimization opportunities
  • Forecasted warehouse inventory and capacity using time series models for warehousing decisions
  • Developed advanced SQL views pulling data from AS/400 and M3 into Excel
  • Shadowed Grower Services Manager to master fruit procurement, reporting, and grower equity
  • Designing linear programming model to optimize fruit receiving by matching suppliers to facilities
🎯 Optimizing transportation efficiency & reducing costs
June 2024 — August 2024
Cyber Data Engineer Intern
United Airlines, Chicago, IL
  • Built and automated data pipeline in Spark, sourcing from a data lake
  • Reduced AWS Glue job run time by 80%
  • Enabled 10x improvement in scalability
  • Developed cloud-based cybersecurity dashboard for big data visualization
⚡ 80% faster · 10x more scalable
April 2024 — September 2024
Data Science Consultant
Gemstone Lights, Calgary, AB
  • Analyzed unstructured IoT controller data to identify device connection patterns
  • Saved multiple hours weekly in IoT testing by flagging devices needing adjustments
  • Built visualization web app using Canada/US census data with filters to target new markets
📊 Hours saved weekly · New markets identified
May 2023 — September 2023
Data Science Intern
WPA Intelligence, Washington D.C.
  • Built ML models predicting voter behaviors with 90%+ accuracy
  • Aided strategies for 25 campaigns across 22 states
  • Created R script automating campaign report generation—increasing productivity 97%
  • Web scraped and standardized 100,000+ rows of data
  • Summarized 50,000 articles for campaign feedback
🏆 1st Place I-Hack Hackathon · 25 campaigns · 22 states

What I've Built

Agentic systems, ML pipelines, and tools that verify their own success.

Compa (FlipAssist)
Active

Real-time arbitrage engine for Facebook Marketplace. LLM-powered title normalization, eBay sold comp fetching, query refinement, and profit calculation—all in a Chrome extension with a docked rail UI.

TypeScript Fastify OpenAI eBay API Chrome MV3
Verifiable Outcome
Real-time profit signals
BUY/MAYBE/SKIP decisions backed by sold comps
WatchFlipper
Completed

Automated watch arbitrage system. Scrapes watch listings, identifies undervalued pieces using ML price estimation, and alerts on profitable opportunities.

Python BeautifulSoup ML
SimpleField
Active

TypeScript-based field data collection and analysis tool. Streamlines agricultural data gathering with structured forms and real-time validation.

TypeScript Data Collection
Dynamic Time Series
Completed

Advanced time series analysis toolkit with dynamic clustering. Processes temporal data to identify patterns, anomalies, and forecasting signals.

Python Time Series Clustering
Ohio Elect Political Consulting
Completed

Analyzed 8M+ rows of voter data, applied ML to predict party affiliation and turnout. Earned 1st Place at 2023 I-Hack Hackathon.

Python ML Political Data
Verifiable Outcome
🏆 1st Place
2023 I-Hack Hackathon winner
OSHA Document Clustering
Completed

Time Series Document Clustering on 5,000 OSHA documents. Used NLP and unique clustering to track topic trends, emergence, and seasonal significance.

Python NLP Clustering
Gemstone Census Map
Completed

Interactive visualization web app using Canada and US census data. Filters and maps to identify and target new markets for business expansion.

Python Dash Geospatial
Jacobian Neural Network
Completed

Calculus 3 learning project implementing neural networks with Jacobian matrices. Deep understanding of backpropagation through matrix calculus.

Python Calculus Neural Networks

The Toolkit

Languages, frameworks, and domains I've mastered through real projects.

🐍 Python Ecosystem

Pandas NumPy Scikit-Learn PySpark Polars NLTK TensorFlow BeautifulSoup4 Streamlit Geopandas

📊 R & Visualization

Tidyverse dplyr tidyr stringr tidymodels lubridate purrr Mosaic Quarto

☁️ Cloud & Data

AWS (Glue, S3, EC2) Databricks Spark SQL AS/400 Infor M3 ERP Infor Birst

🤖 AI & ML

LLM Integration Agentic Systems Time Series Clustering Classification NLP Linear Programming OpenAI API

🌐 Web & Extension

TypeScript Fastify Chrome Extensions HTML/CSS Plotly Dash Tableau AWS Quicksight

🛠️ Tools & Methods

Git/GitHub ArcGIS Microsoft Excel C# Fluent Spanish Agile

Thoughts & Learnings

Documenting the journey from data to decisions.

2026 — Coming Soon
Building Compa: Why I Rejected "Vibe Coding" for Agentic Architecture
Most AI-powered tools are wrappers. I wanted to build something that reasons. Here's how I architected a real-time arbitrage engine where LLMs don't just generate—they decide.
AI Architecture Agentic Systems
2026 — Coming Soon
From 80% Faster to 10x Scalable: Lessons from United Airlines
How I rebuilt a cybersecurity data pipeline in Spark, reduced AWS Glue costs, and learned that premature optimization isn't evil—it's necessary at scale.
Spark AWS Data Engineering
2026 — Coming Soon
Linear Programming in the Orchard: Optimizing Fruit Supply Chains
Applying operations research to agriculture. How I'm using LP models to match suppliers to facilities and why the real world is messier than any textbook.
Optimization Supply Chain Agriculture

Let's Build Something Intelligent

I'm looking for AI/ML Engineering or Data Science roles where I can turn complex problems into verifiable solutions. If that sounds like your team, let's talk.