Check out the roadmap to get a better idea of the status of individual projects within my portfolio. Generally speaking, if its in "review", it means, I'm reasonably confident that it's worth showing a potential employer, friend, business partner, and would LOVE some feedback on from any of the people who have cloned this repo (i do watch my insights). Otherwise, you can assume its either a work in progress, used for demonstration purposes or I'm unsure if I want to make it public for legal/financial reasons
This repository is a collection of Jupyter Notebooks showcasing my work in Data Science, Quantitative Finance, Machine Learning, and AI. I strive to take unique and often experimental approaches to problems. As a result, some files exist as outlines, either because I haven't finished them or because it would be incredibly unwise to publish them. I have no obligation to clarify which is which, though the roadmap may give you some indication of that.
This portfolio represents only a small fraction of the code I'm willing to publish and falls into three categories:
- General Data Science: Contains generic, mostly utility scripts and projects in Python.
- Social Science: Primarily concerned with social and economic data and analysis. Link
- Quantitative Finance: Comprises the bulk of the portfolio, including anything I'm willing to publish that can save or make money. Link
If you want to learn a little about me, know that I grew up illiterate and largely unable to speak my first language until I was 12. So, I don't have the luxury of thinking I'm any smarter than anyone else. However, I developed a passion for learning anything and everything that interested me. I graduated high school with honors, went to college and studied applied Sociology, published a paper under a pen name, audited enough grad courses to have my master's degree, but didn't want to take the GRE.
I then found startup weekend, learned coding, and I've been working on and off as a full-stack JavaScript developer for over 10 years. I've been both a TA and developed and taught full-stack curriculum for high schoolers at WestMec. I've also written extensively on the subject on Medium and manage 3-4 small publications there.
After COVID, like most people, I faced unemployment and received stimulus money. However, unlike many, I had no reason not to finally write code to manage that money (which I've been thinking about since graduating). I started with pinescript and thinkscript, just reverse engineering code and mixing and matching things, meticulously backtesting, paper trading, and finally trading with real money. Spoiler, I'm still trading and living off that money.
After GPT came out, I spent a lot of time training custom financial models like GPT and other local models, which I use extensively for coding Python. I've published a repo on the GPT for my initial prompts. However, the bulk of the files my version is trained on are no longer public. The ones there were used to train it export partial or, in some cases, full Jupyter Notebooks.
A big reason for this is I spent a lot of time on the rights to train it on high-level math and other subjects I'm not an expert in and frankly don't have the time or interest to learn. So, this custom GPT and other models help fill a very valuable knowledge gap typically reserved for PhD students.
It's my personal belief that any well managed Quantitative Portfolio, should be ACTIVELY managed, semi-automatously and think in term of Recursive Investing. Moreover, it should be well diversified with correlated assets and uncorrected assets, and attempt to seek both income and growth regardless of market conditions. Each segment, using machine learning, forecasting and automated risk management is absolutely essential in this approach. Understanding this, is what makes the difference between millions and billions...
For example,
- Bonds should be held to produce currencies, which are held and borrowed against for living expenses.
- Currencies, should be weighted based on the individuals home nations, the nations they do business with, and should be multiplied based on macro economics, and arbitrage.
- Dividends, should NOT be DRIPPED, they should be used to cover the risk of margin trades, or reinvested into growth, so as to outpace, the potential decay.
- Growth Stocks, should fall into 2 categories, those with divided and those that are speculative.
- Options, should be sold, rather than bought. We're not gamblers, we're investors... Sell to the degens and make the market, rather than being played by it.
- Real Estate, should be invested in by borrowing against a well managed portfolio of the above assets. It goes without saying, that real estate is great for both income and growth. More importantly, however, real estate is ultimately about investing in communities, people and the commons. Tax liens, are a third way to invest into real estate, which is far too commonly overlooked. Finally, when a property is nearing paid off, or when interest rates have flipped into a decline, the loans should be refinanced to invest into the assets mentioned above. In this final segment, we start all over again and complete the recursive loop.
If you're interested in hiring me, you may message me on my LinkedIn or checkout my leetcode profile, however, know that if you are cold messaging me, I will not do a technical interview. I've been writing code and solving complex multidisciplinary problems for 10+ years and writing code to trade is way more profitable and practical use of my time, than going through 3-5 rounds of unpaid coding projects specifically designed to gaslight people and cause imposter syndrome. I wish the best and mean no disrespect for you or your company but at the end of the day my dignity, my time, my mental health and my son will always be more important than any salary or equity you can offer me.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent