About

{about me_}

I’m George Sánchez, a research analyst and first-generation college graduate, born and raised in the South Bronx, NYC. My work has focused on developing systems for investment analysis, where I combine traditional financial acumen with data science to extract insights from structured and unstructured data, automate investment decisions, and ultimately explore what makes markets tick—in both private and public markets.

*** [optional] click on right-side dropdown carets for additional detail

Education_

I graduated from Duke (’22) with a B.S. in Financial Economics. My studies covered topics ranging from Mathematical Economics, to Asset Pricing & Derivatives, to Behavioral Finance, providing me with the foundation needed to excel outside the classroom; I was a consistent medalist in Duke’s case competitions, and translated this success into multiple internships with Rothschild & Co and Ulu Ventures. In search of greater challenges, I joined Ulu Ventures’ internal quant fund, Blockspeed, as employee #1, taking the role of research analyst.

I have known that I wanted to do this kind of work since 6th grade, when our class participated in the National Stock Market Game. The idea that people used math and computers to form predictions which made money as they slept was captivating, especially to a nerdy kid whose family otherwise had to fight for every penny. Fast forward to 2022 and I was graduating from Duke University with a Bachelor of Science in Financial Economics.

There, my studies covered topics ranging from Mathematical Economics, to Asset Pricing & Derivatives, to Behavioral Finance. I also managed to distinguish myself through Duke’s finance case competitions, becoming the only student at the university to place top-3 several times. I then parlayed this success into multiple summer + co-op internships with Rothschild & Co (the only such IB intern at the firm) and Ulu Ventures (their first ever undergrad intern).

These experiences, though not as technical as my later work, were crucial, as they provided me with a strong foundation in economic theory and financial analysis. Perhaps more importantly though, they fueled my desire to further pursue the kind of deep statistical modeling, research, and data-driven storytelling that I amorphously envisioned as a teenager. As luck would have it, during my final semester at Duke, Ulu extended me an offer to join their internal systematic trading fund, Blockspeed, as employee #1.

Building a hedge fund from scratch_

At Blockspeed, we developed trading strategies for digital asset markets based on how productive each asset was (according to our domain-specific KPIs). My research doubled our tradable signals, and after insisting on taking over infrastructure development, I single-handedly improved our E2E execution times by 100x. Our model outperformed benchmarks in live trading, achieving +8% returns vs benchmark’s -4%, with lower drawdowns.

As the first hire at Blockspeed, I worked directly alongside our CIO Stephen Malinak—who previously led quant R&D teams at ADIA, Thomson Reuters, and StarMine—to develop “quantamental” strategies for inefficient digital asset markets.

Our model combined fundamental metrics with momentum indicators, tailored to the unique characteristics and data-generating mechanisms of digital assets. My research, in particular, focused on understanding drivers of value across these projects, deriving signals/features to enhance the predictive power of the model. These signals combined metrics based on user activity, developer engagement, supply activity, and more.

Although I was initially hired for financial research, I soon raised my hand and took over development and optimization of our data+research infrastructure, improving execution times by 10-100x through a combination of NumPy, Numba, Async, and Ray. To support the research process, I also built comprehensive tools for scenario analysis and stress testing. These tools employed data-shuffling, Monte Carlo bootstrapping, and general hyperparameter jittering to minimize the risk of overfitting.

During my tenure, our strategies demonstrated strong performance in live trading, outperforming our benchmark in both absolute returns (+8% vs -4%) and max-drawdown (23% vs 32%).

Bringing data science to venture_

With Ulu Ventures, I’ve developed tools to streamline various investment and operational processes. These included (1) investment memo generation, (2) semantic search over legal documents, (3) targeted investor outreach, and (4) automated financial oversight of portfolio companies. Many addressed operational blind-spots, others were universally praised time-savers, but most importantly, they all resolved the pain-points they were designed to address.

At Ulu Ventures, I’ve worked closely with the Partners to devise tools and processes that maximize the company’s vast stores of structured and unstructured data. We were primarily focused on addressing pain-points related to repetitive due-diligence questionnaires, proactive portfolio oversight, and cumbersome investor data.

To prevent time-consuming meeting delays, I created an AI system that autonomously generates investment memos by synthesizing meeting transcripts, company notes, and web research. This ensures all team members are well-informed about potential investments, saving hours worth of time that was otherwise being wasted catching people up on companies mid-meeting.

To streamline due diligence, I built a semantic search tool that quickly retrieves relevant past responses to questionnaires, significantly reducing response time and ensuring consistency in our communications. For more efficient investor outreach during fundraising, I developed a contact-enrichment system that identifies and prioritizes high-value investors, even with incomplete initial data. This system uses web sources, PitchBook, and LinkedIn data to find the most promising contacts for targeted outreach.

Lastly, I improved our portfolio management process by creating a system integrated with StandardMetrics, a service the firm uses to request monthly financials from our portfolio companies. The system monitors the data quality of these submissions, checks the general financial health of our portfolio companies, identifies concerning trends, and automatically notifies us and the founders via email when conditions for action are met. This proactive approach allows us to better support our portfolio companies and maintain reliable transparent reporting to our own investors.

These tools and systems significantly improved our operational efficiency and decision-making processes, allowing us to evaluate more deals, conduct thorough due diligence, and provide better support to our portfolio companies.

Activities and interests_

When I’m not deep in code or financial models, you can find me:

On a walk with my high school sweetheart, Greer, as we do our best impression of an elderly couple.
Watching and over-analyzing a Formula-1 race or NBA game, likely loudly and out of my seat.
Playing Gran Turismo, where I compete in sim-racing tournaments and design cars for virtual photoshoots.

Fun facts_

My birthday is on 2-Pi (aka tau) day: June 28.

Our family has gone on a road-trip from NYC to SF and back. I nearly slipped off a cliff in Utah, but thankfully I’m still here to write this fun-fact.

I couldn’t beat my 55+ year-old mom in an arm-wrestle until I was nearly an adult…

…because my mom immigrated to the US from Dominican Republic specifically to advance her boxing and power-lifting career.

According to Gran Turismo, I am on average the quickest driver in all of NYC, though I’ve previously peaked at #1 in the entire state.