From the driver's seat to production AI for federal clients. Scroll — I'll take you through it the way it happened.
The discipline it took to finish a degree that way is the same discipline I bring to production systems now.
But the car was only half the story.
The other half was written in code — and it started with two words.
First CS course at Florida International University, taken between delivery runs. Two words on a screen — and the first time a machine did what I told it to.
Boeing had just acquired a Miami airplane-parts company and needed its databases in sync for federal-contractor compliance. Records were full of nicknames — so I built fuzzy-logic matching to statistically resolve each one to a legal government name and link profiles across systems.
On the Mastrena II espresso platform, I used ETL and data analysis to find the settings that minimized downtime — then, with two MIT graduate students, built a predictive model that flagged failing machines and opened a maintenance ticket before the breakdown.
Mid-acquisition, I upgraded Citrix's existing machine-learning models — hyperparameter tuning and new evaluation metrics to measurably improve on what was already running in production.
A personal project caught the attention of a graduate professor, who became a voice for my work — a $6,000 grant, a place in FIU's SOLID Lab, and a public dashboard forecasting COVID-19 deaths and infection rates that the university recognized as one of its leading projects.
Graduated — self-funded, every semester — and returned to the team I'd interned on. Shipped demand forecasting across 11,000+ stores with $79M in A/B-validated annual savings, plus drift detection that caught degradation before it hit operations.
AI/ML Engineer building governed GenAI and ML systems for federal enterprise clients — AUC-ROC from 0.82 to 0.91, audit compliance from 72% to 98%, manual review down 60%+.
In my senior year, one of my personal projects caught the attention of a graduate-level professor. He saw the work — and decided to become a voice for it.
With his backing I won a $6,000 research grant and became an Undergraduate Research Fellow (NSF REU, Spring–Fall 2020) in FIU's SOLID Lab. Masks on, models running — a public dashboard forecasting COVID-19 deaths and infection rates.
The university recognized it as one of its leading projects — and it became a lead-author, peer-reviewed paper in Patterns (Cell Press). That's when I understood what a platform could do.
M. Hyman, C. Mark, A. Imteaj, H. Ghiaie, S. Rezapour, A. M. Sadri, M. H. Amini. Patterns, 2(8): 100315, 2021.
The professor brought me into FIU's SOLID Lab — Sustainability, Optimization, and Learning for InterDependent networks — where I built the pipelines, the model, and the public dashboard behind the paper.
After years of shipping other people's systems, I bet on my own. DeepBlueAlpha is a full-stack quantitative data platform, founded and built solo — nights and weekends, the same grind that paid for college, pointed at my own thing.
Python pipelines ingest and score SEC EDGAR filings for 5,000+ equities with automated data-quality validation — Piotroski F-Score, Altman Z-Score, Beneish M-Score, and options models (Black-Scholes, Heston, Bates) — evaluated against historical benchmarks.
It started as another after-hours personal project: one dev, one dark room. Then I stumbled onto something — instead of hand-writing a worker's rules, you could clone the worker.
Capture the attributes of an actual worker — their judgment, patterns, and decision-making — and reproduce them as an AI/ML system built in that person's image. Any role's expertise, cloned and deployed at scale. I've been building on it ever since.
I write and speak about what it actually takes to put machine learning into production — and about widening the door for people building their way up the hard way.
Essays on production ML, MLOps, and trustworthy AI. Coming soon.
Available for talks and panels on shipping ML at scale. Get in touch below.
Especially for first-gen and self-funded students breaking into data science.