ML invoice reader
A data-scraping and machine-learning pipeline that reads vendor invoices and matches them against accruals automatically — built with machine learning before ChatGPT existed.
Operator-Engineer
I find the broken process. Then I build the fix.
Not a consultant with a deck — a one-man shop. I identify bottlenecks, understand the business logic, and build and ship full-stack software solutions — agentic workflows, databases, and custom frontends. Fifteen years across aerospace, global logistics, and software.
B.Eng (ECE) · MBA · CFA Level I
Three layers, one head.
Most teams need a translator between the strategist, the analyst, and the engineer. I hold all three at once — nothing gets lost in the handoff, because there is no handoff.
Engineering
B.Eng in Electrical & Computer Engineering. Former aerospace engineer. I debug a business the way I debug a circuit — trace the signal, find the fault, fix the root cause.
Finance
MBA and CFA Level I. I speak fluent P&L — working capital, pricing, margin, forecasting — so the software I build is aimed at the number that actually matters.
Full-stack software
Self-taught, Python through TypeScript, SvelteKit, Supabase, and Cloudflare. I design, build, ship, and operate production platforms end to end. The work speaks for itself.
The proof
Seven years inside a billion-dollar operation.
I spent seven years as a data scientist at the Quick Group of Companies — Kuehne+Nagel’s time-critical air-freight division, moving healthcare, pharma, and aerospace shipments. The back office was drowning in manual work. I didn’t write recommendations. I wrote the systems that replaced the work.
~30
full-time roles automated across back-office departments
90 → <30
days payable — direct working-capital impact of my Python systems
7 yrs
embedded inside a billion-dollar P&L
Dynamic billing algorithm
A system that autonomously rates and adjusts billing for the company’s largest accounts, replacing slow, error-prone manual work.
C-suite dashboards
Real-time operational and financial KPIs, wired straight to the source data, for executive decision-making.
How I work.
The strategy, the financial model, and the codebase come from one person.
No hand-offs. No translation loss between the business case and the implementation. I sit with the operation, learn where it actually leaks money, model the fix, and then write the software that runs it — in production, owned end to end.
That’s the whole point: the person who understands the P&L is the same person shipping the code. It moves faster, and it aims at the right number.