Projects

PersonaLedger is a system and dataset for generating high-quality, synthetic financial data. The dataset is entirely created using self-enforcing LLM agents, and was developed as a joint research effort between Capital One and UMD. We have benchmarked it by using it the dataset to predict customers who will borrow beyond their spending capacity, and to identify identity theft in transaction sequences. Code and datasets are found in therepo.

The generation system works in a self-correcting loop: an LLM generates user “personas” along with transaction data matching the persona, then agentic and code validators enforce financial rules, and invalid generations are fed back as constraints to improve future generation. You can find the PersonaLedger research paperhere.

I wanted to learn the math behind neural nets so I wrote a simple digit classifier using just linear algebra. The classifier recognizes digits from theMNIST Datasetand implements a forward and backpropogation algorithm to train itself.

This is the website you are currently reading. I wanted to gain experience coding websites and apps, and build a space to interact with the world.

This website is built usingNext.js 14,tailwind css, and deployed withVercel.

This is a simple web scraper I built. This was one of my first coding projects, and I hoped to achieve three things:

In Chicago there is a bar calledTin Lizziesthat hosts turtle races. Drunk patrons can bet and cheer on much loved and definitely not exploited turtles that meander around an arena. Considering how popular this was in the local neighborhood, my good friendSam Royalland I thought this would lend itself to a great gambling app.

I was very interested in brownian motion, and b-splines in computer animation, so I focused on procedurally generating the smooth random path the turtle would take. You can find a demo of our app in thegithub. Happy gambling!