About

Fundamental by training. Quantitative by method. Technical by habit.

Three disciplines that are usually hired separately, applied to the same problem: what is this asset worth, how do I know, and can I prove it again next month without me in the room.

The three are not a career change. They are one method.

I learned markets from the balance sheet outward: value investing at Alabama, then a year of private equity diligence building LBO models and reading filings until the numbers stopped surprising me. That training is why I start with where cash comes from and what someone is paying for it, rather than with what a chart has done lately.

The quantitative layer came from a specific frustration. A good thesis could not be tested, and an untestable thesis is a story with a price target attached. So I learned to write the thesis as a rule, test the rule across regimes, deflate the result for how many things I tried, and size it so that being wrong was survivable.

The technical layer came next, because doing that by hand does not scale past three ideas. I build the data layer, the backtest engine and the automation myself — Python, FastAPI, C++ where speed matters — so that a result can be reproduced by someone who is not me, on a machine that is not mine.

Each layer makes the other two more useful. The fundamental work says which questions are worth testing. The quantitative work says whether the answer survives contact with data. The engineering says whether any of it can be trusted twice.

02 / The overlap

Why the combination is worth paying for.

A fundamental analyst without code produces one memo a week and cannot re-run it. A quant without accounting builds signals on numbers they do not understand the accounting behind. An engineer without markets ships infrastructure nobody asked for. The value is not in any one of those, it is in refusing the trade-off.

Concretely: I can read a 10-K and tell you which line item is doing the work, then write the screen that finds the same pattern across a universe, then build the system that runs it nightly and reports when it stops behaving. The same person did all three, which is why the outputs agree with each other.

That is also why the failures are published. A candidate who only shows wins is asking you to trust the highlight reel; the record here includes a meta-labelling model with no out-of-sample skill (AUC 0.51), barrier exits that destroyed the trend premium, and a strategy family rejected on cost. Those are the parts that show the process is real.

03 / Experience

The record.

Nov 2025 — Nov 2026

Junior Research Associate

Potomac Fund Management · Bethesda, MD

  • Originated and developed a systematic multi-asset allocation strategy, translating an investment thesis into a rules-based engine and advancing it from research and backtesting to live execution with client capital.
  • Researched, coded and backtested systematic strategies across investment areas beyond commodities, using custom indicator scorecards and dynamic risk-allocation models to evaluate opportunities.
  • Evaluated performance across market regimes and built optimization and stress-testing frameworks that account for slippage and structural market shifts before deployment.
  • Authored strategy narratives, performance analyses and methodology documents for prospective institutional partners and third-party platforms.
  • Architected and deployed a proprietary multi-agent AI platform using Python and FastAPI services, with an internal UI for research workflows, automating roughly 90% of the research process from idea generation through testing, coding, iteration and documentation.
  • Built the research execution layer and data bridges (originally AmiBroker/AFL, now a C++/Python engine), including OCR and document-parsing pipelines for unstructured institutional data.
Systematic
strategy
Automation
Jul 2024 — May 2025

Private Equity & Market Research Intern

Maximus Capital Partners · McLean, VA

  • Supported a $50M investment decision by analyzing financial statements and market trends during due diligence.
  • Built a comprehensive LBO model in Excel to project portfolio performance for a target acquisition and presented findings to senior stakeholders.
  • Delivered investment analysis reports synthesizing research for cross-functional teams.
Due diligence
LBO modeling
Valuation
Aug 2024 — Dec 2024

Graduate Teaching Assistant

University of Alabama · Tuscaloosa, AL

  • Supported the undergraduate real estate program with feedback on valuation and financial modeling assignments; helped integrate case studies and current market trends into coursework.
  • Improved grading efficiency by 20% by implementing standardized rubrics in a high-volume course.
Valuation
Financial
modeling

04 / Education

Formal training.

Aug 2024 — Aug 2025

Master of Science, Finance

University of Alabama · Tuscaloosa, AL

Focus: corporate finance and financial planning & analysis. Graduate teaching assistantship in the real estate program.

M.S.
Finance
Aug 2021 — May 2024

Bachelor of Science, Finance

University of Alabama · Tuscaloosa, AL

Focus: value investing and quantitative finance.

B.S.
Finance

05 / Capabilities

What I actually work with.

Grouped by how the work gets done. Every item here has been used on a live project, a deployed strategy or a published piece.

01 Fundamental analysis

  • Financial statement analysis
  • Equity research and valuation
  • DCF and LBO modeling
  • Comparable company and transaction analysis
  • Due diligence and market research
  • Normalized and mid-cycle earnings
  • FP&A and performance analysis
  • Investment memo writing

02 Quantitative research

  • Systematic strategy design and rule specification
  • Backtesting and walk-forward validation
  • Deflated Sharpe and multiple-testing adjustment
  • Regime analysis and stress testing
  • Dynamic risk allocation and position sizing
  • Indicator scorecards and composite signals
  • Volatility targeting
  • Options and defined-risk structures
  • Multi-asset and commodity strategies

03 Software & data

  • Python, pandas, NumPy
  • SQL and data modeling
  • C++ for performance-critical engines
  • FastAPI microservices
  • Next.js and React
  • QuantConnect / LEAN
  • AmiBroker Formula Language (AFL)
  • MetaTrader 4 and 5
  • Data pipelines and artifact stores

04 Systems & automation

  • Multi-agent system orchestration
  • LLM gateways and structured-output validation
  • Sandboxed code execution
  • OCR and document parsing
  • Win32 API and application bridges
  • Reporting automation (Excel, Power BI, pptx)
  • CI regression suites for research engines

06 / Where I fit

Two lanes, one method.

Investment research and portfolio analytics. I can own a coverage list, build the model, form a view and defend it in writing — or sit with a portfolio analytics team and build the attribution, risk and reporting that shows what is actually driving the book.

Systematic research and research engineering. I can specify a signal, build the backtest, run the walk-forward validation, and document the failure modes so a risk committee can inspect the process rather than trust a pitch.

The evidence sits on this site: four company reports, eight backtested strategies with their failures included, and the architecture for the platform that would run them. None of it is a claim about returns; all of it is a claim about process, which is the thing that transfers between seats.

Open to full-time roles across the United States and to remote teams.

[email protected]