Data Scientist & ML Engineer · MSBA @ University of Minnesota · Ex-Tredence Analytics
LinkedIn · 📧 just4dec@gmail.com
I build things that work — end-to-end data pipelines, autonomous AI agents, causal studies, and trading systems. Over 3+ years across CPG and finance, I've shipped forecasting models for Fortune 100 companies, designed streaming data warehouses, and explained why a $M campaign flopped in Thailand.
Currently finishing my Master's in Business Analytics at UMN, where I spend most of my time applying causal inference to messy real-world problems or building agentic systems that replace manual workflows.
Outside the terminal: boxing, lifting, and anything that requires discipline and reps.
"In God we trust. All others must bring data." — W. Edwards Deming
| Project | What it does |
|---|---|
| Project SPEAR | Autonomous B2B outreach engine — LangGraph + ChromaDB + GPT-4.1. Replaces a 5-person marketing team, 10× volume at 20% cost |
| HermesAI | Real-time audio streaming assistant — live sentiment, intent detection, and next-best-response during active calls |
| Project | What it does |
|---|---|
| VCPBot | Fully automated Minervini VCP swing trading bot — 7-phase pipeline, Alpaca bracket orders, deployed on Oracle Cloud |
| Optiver Trading | GRU + Optuna solution for Kaggle's Optiver competition — forecasts auction-close price movements from order book data |
| Project | What it does |
|---|---|
| MinneMUDAC 2025 🥈 | Predicted mentorship match longevity for Big Brothers Big Sisters — NLP on case notes + structured ML pipeline. 2nd place overall |
| TABOT Impact Study 🥈 | Causal analysis of a TA chatbot's effect on student grades — DiD + regression. 2nd place, Analytics for Good Hackathon 2025 |
| NIL Policy & NCAA Athletes 🏆 | DiD study on NIL policy's impact on basketball performance, using Canadian players as a natural control group. 1st place case comp |
| Project | What it does |
|---|---|
| Freemium Conversion | Classification model on 41K users (3.7% base rate) — targeting strategy achieved 219% projected ROI |
| Cats vs Dogs | CNN + transfer learning for Kaggle's Dogs vs Cats competition — progressive model refinement tracked via log loss |
Let's build something that actually ships. 🚀