@knightfallShashy

I'm building a finance calculator, where my models compound growth over time and checks retirement projections against the 4% rule. I built 4 different calculation based tabs. Additionally, I used streamlit to launch my website. I used only python for this project.

Investie: a full-stack AI-powered investment research platform (like a lightweight Bloomberg Terminal + Yahoo Finance + ChatGPT combo). Built with Reflex (Python frontend), FastAPI (backend), PostgreSQL, and the Anthropic API for AI-generated company summaries, bull/bear analysis, and grounded Q&A. Users can research stocks, track portfolios, run DCF valuations, and get AI-powered investment insights, all through a modern dashboard.

A FICO score quantization engine that maps continuous credit scores (300–850) into optimal buckets using dynamic programming to maximize log-likelihood of mortgage default prediction.

A credit risk model that predicts the probability of default (PD) for personal loan borrowers using logistic regression, and estimates the expected financial loss per loan. Built as part of the JPMorgan Chase Quantitative Research Virtual Experience on Forage.

A Python pricing model for natural gas storage contracts that calculates the fair value of a buy-store-sell trade based on injection/withdrawal dates, historical price data, and storage costs. Built as part of the JPMorgan Chase Quantitative Research Virtual Experience.

This project analyzes historical natural gas prices and builds a function to estimate the price of natural gas for any given date, including dates beyond the available dataset. Type any date between & after October 2020 and September 2025 and press Enter.