I'm currently building quant trading strategies at Edgehog Trading.
I'm based in Chicago via NYC, Boston, DC, and Medellín.
I studied Math with Computer Science at MIT, and have been using computers to solve interesting problems ever since.
I've worked in finance, tech, consulting, and startups.
I've also been a college athlete, a club DJ, and a competitive chess player.
I'm intrigued by languages, music, history, philosophy, travel, design, playing sports, good conversations, and learning things both useful and frivolous.
At Edgehog, we are building the next generation of electronic option market making strategies. I build trading strategies and real-time tools for traders and developers in support of this mission. We work with some very committed and talented people, and we're usually looking for more! If you're interested, please don't hesitate to reach out.
To inform investments, I built multiple-input (text + numerical features) deep learning models to predict the outcomes of lawsuits. In addition, I architected and maintained Patronus's end-to-end tech stack.
I researched, created, and managed new and existing trading strategies. In addition, I mentored and managed newer traders, controlled desk risk, and spec'ed and prioritized software and strategy improvements.
I provided predictive analytics consulting for two different clients. In addition, I presented studies' results, built relationships, and trained clients to use proprietary software.
I built a customized news parsing web app which aggregates actionable news from hundreds of online sources in real-time.
I developed a data oversight web app to monitor and clarify Amazon Web Services expenses for a large data firm.
I built parts of the GUI for a revolutionary web-based Digital Audio Workstation.
Whether it's predicting litigation outcomes, trading financial markets, or determining business strategy―a high degree of uncertainty is inherent in any complex project. The uncertainty doesn't end when the model stops training, when P&L is realized, or when a final decision is made. This uncertainty, though terrifying, is what makes projects interesting and worthwhile in the first place. Managing it requires healthy doses of pragmatism, paranoia, and a grounded knowledge of the mathematical and computational tools available.
I've built models on a variety of data, algorithms, and contexts.
From deep learning neural networks, SVM, and Random Forest to
linear regressions, KNN, and PCA,
I've used and developed an understanding of a broad range of model types.
In terms of data itself, I've worked most with numerical time series,
but I also have experience within other domains like natural language processing and computer vision.
My data science toolbox includes most of the usual suspects: Python (TensorFlow, Keras, Pandas, NumPy, sk-learn), R, and SQL, to name a few.
Because my work often requires me to step outside of a Jupyter notebook,
an understanding and appreciation of good software engineering principles is essential.
In addition to data science, I have experience across the stack
and have built everything from web app UIs to backend architectures.
Technologies like Docker, Git, and AWS have proven their worth time and time again in my projects.
My career in quantitative trading has spanned multiple desks, asset types, and strategies.
It's made me comfortable working in an environment that's fast-paced and high-stakes, with a low signal-to-noise ratio.
It's also given me a deep understanding of the complexities of trading particulars
like market microstructure, exchange rules, and margining algorithms.
I've worked for early-stage startups both as an employee and as a co-founder.
In addition to engineering in a startup context,
I also have experience conducting customer interviews,
doing market research, and making hard business decisions under extreme uncertainty.
It's taught me valuable lessons in lean development, customer discovery,
and building things that people actually want.
Web development is not a core competency, but I have experience in it all the same.
Specifically, I've worked with tools like Flask, D3.js, Angular, Javascript, and HTML+CSS.
I've built web front ends for a variety of use-cases,
and every once in a while it comes in handy for data science purposes.
Plus, it's pretty fun.