About

I studied finance and marketing at Purdue, with a concentration in data analysis. My first jobs were finance jobs: forecasting and variance analysis on a $7.2B defense program at Raytheon, and investment modeling at a real estate firm I helped start.

In every one of those jobs I ended up building the model, then the pipeline behind it, then the tool that meant nobody had to rebuild the model by hand. Somewhere along the way the building became the job. In between, I co-founded a LiDAR startup and trained perception models for it in 2021.

Now I'm an AI data scientist at Caterpillar, building agent systems, retrieval, and fine-tuned models for the Finance and Investor Relations teams. I also co-founded OFFBEND, an independent studio in Dallas, where I build our software: a company agent platform, an AI acting coach, and a screenwriting app.

The finance background isn't a detour. It's why I'm useful. I know what a tick-and-tie is, why one false alarm can cost a tool its credibility, and what it takes for a number to be right. So I build AI for work where mistakes are expensive, and I measure it before I trust it.

Experience

1. CaterpillarAI Data Scientist

   2026 – PresentIrving, TX

   I own an internal agent platform for Caterpillar's core enterprise Finance and Investor Relations teams, end to end.

   • Agent platform used by the 50-person core Finance and IR teams, with expansion planned into Legal and Communications.
   • Tariff classification with a fine-tuned ModernBERT plus an agentic verifier: measured accuracy went from 7% to over 80% across a 1M+ part backlog.
   • Custom RAG engine that catches 99% of document errors at a <0.5% false-alarm rate.
   • LangGraph multi-agent workflows that save each user 10–20 hours every earnings cycle.
   Agents for earnings season →Tariff classification at scale →

2. CaterpillarFinance Data Platform Engineer

   2024 – 2026Irving, TX

   I built the data and AI tooling finance runs on, from legacy ERP pipelines to a conversational SQL tool.

   • GenAI text-to-SQL tool used by several thousand finance employees a month.
   • A separate Next.js training platform for Caterpillar's OneStream rollout, used by 3,000+ employees.
   • Batch and near-real-time pipelines unifying 30+ SAP, Oracle, IBM DB2, and Teradata sources into Snowflake.
   Finance data platform →

3. OFFBENDCo-Founder

   2024 – PresentDallas, TX

   An independent film studio that builds its own tools. I own architecture, infrastructure, and releases.

   • Oppie, a private LangGraph agent platform with policy-based authorization, approval gates, and sandboxed code sessions.
   • Sides, an AI audition coach on an event-driven AWS pipeline.
   • First Page, a local-first screenwriting app for web and native macOS.
   • Scene, a native iOS film and TV journal with a hybrid recommender, now in beta.
   • Led a distributed engineering and creative team across three shipped products.
   Oppie →Sides →First Page →Scene →offbend.com and Connect →

4. RaytheonProgram Financial Analyst

   2023 – 2024Remote

   Finance point of contact for lifecycle services on the $7.2B Javelin Joint Venture.

   • Built Power BI, SQL, and SAP reporting for forecasting, variance analysis, and executive visibility across 200+ contracts.
   • Line-item analysis, actuals reporting, and EAC support between finance operations and program execution.

5. TyriadCo-Founder and CEO

   2021 – 2023Indianapolis, IN

   An automotive startup building full-stack FMCW LiDAR and perception for vehicle autonomy.

   • Sensor architecture built around a single eye-safe 1550nm laser and dual-axis scanning mirrors.
   • Trained semantic classification models for scene understanding in 2021.
   • Real-time data processing and device management on AWS Lambda and IoT Core.
   Tyriad →

6. HNLCFounding Engineer

   2020 – 2024Indianapolis, IN

   A real estate investment firm that grew into software for sourcing, appraising, and analyzing deals.

   • Built Python, SQL, Databricks, and Spark pipelines for investment modeling across thousands of real estate assets.
   • Shipped Tableau and Power BI products for sourcing, appraisal, and market analysis.

Education

Skills

AI

LangGraph, RAG, hybrid retrieval, agent orchestration, evals, tool calling, fine-tuning, agent governance

Cloud

Azure, AWS, Cloudflare, Vercel, Modal, Terraform, ARM, GitHub Actions, Docker

Languages

Python, TypeScript, SQL, Swift

Data

Snowflake, Databricks, Spark, PostgreSQL

Frontend

React, Next.js, SwiftUI

Systems

Distributed, event-driven, and real-time systems, APIs, schema modeling, local-first sync

Get in touch