Quantitative systemsengineer building rigorousinfrastructure for markets.
Validation frameworks, real-time data pipelines, production trading systems, and the knowledge architecture that ties them together.
By the numbers
Built with rigor, measured honestly.
Validated Market Data
Bars Analyzed
Knowledge Base Entries
Strategies Rigorously Tested
Selected work
Four systems. One thesis.
Each project below was built end-to-end and validated against real market data — not toys.
Catching the lies backtests tell
Validation Framework
Combinatorial purged cross-validation with deflated Sharpe correction. Tested 19 strategies on 7 years of real market data. Built to flag false positives before they ever see capital.
From one day to seven years
Market Data Pipeline
Multi-source data infrastructure unifying Databento + IBKR. Discovered and engineered around per-request cost caps. Built reproducible 6.5-year multi-asset history for under $20.
Production-grade infrastructure
Trading Systems
Pluggable strategy abstraction, next-bar-fill backtester, seven-layer risk stack, IBC autostart for live IBKR. The plumbing that turns research into something that can trade unattended.
A second brain for quant finance
QuantVault
9,607 cross-linked notes across eight domains: strategies, math, microstructure, risk, infrastructure, history, behavioral finance, careers. Designed as production software, not a notes folder.
Approach
How I think about building for markets.
Validate before you trust
Backtests lie. Selection bias is everywhere. I build validation infrastructure first — CPCV, deflated Sharpe, multi-year out-of-sample — because the cost of believing a false positive is real money.
Real data only
No synthetic series. No simulated prices. Every system is tested against actual market data from Databento and IBKR. The framework enforces this with runtime guards.
Risk is a layer, not an afterthought
Production trading systems need brick walls, not check boxes. Daily-loss limits, pre-emptive kill switches, EOD force-close, news blackouts — every safeguard explicit, every failure mode considered.
Honest about what doesn't work
I shipped the validation framework that disproved my own strategies. That's a feature, not a bug. The same rigor that flags my false positives is what makes my real edges trustworthy.
About
I sit at the intersection of two rare skills.
Most engineers can't quantify what makes a trading strategy actually work. Most traders can't ship production software. I do both.
For the last year I've been building quantitative trading infrastructure from first principles — a validation harness that catches false positives, a multi-year market data pipeline, a production trading system, and a 9,600-entry knowledge architecture for quant finance research.
What I'm doing next: applying that same systems mindset to fintech problems where the engineering depth matters as much as the math.
Focus
Quant systems
Location
United States
Status
Open to work
Stack
Python, TS, Next.js
Markets
CME futures, ETFs
Tools
IBKR, Databento
Let's talk
Have a system worth building?
Open to fintech engineering, quant research, and freelance projects. Reach out and let's see if there's a fit.