Open Source

Finance & AI tools, built in the open.

Personal research projects at the intersection of quantitative finance, AI systems, and enterprise analytics. Free to use, fork, and run.

github.com/shishaustin

finance-agent

LLM-powered financial analysis agent

Give it a ticker. It calls live market data APIs, reasons over the numbers, and writes a structured analyst memo — thesis, key metrics, risks, and a bottom line. Built on Claude's tool-use API.

Agentic AIPythonClaude APIFinance

github.com/shishaustin

balance-sheet-forecaster

Bank balance sheet forecasting toolkit

ARIMA-based forecasting for deposits, loans, NIM, and NII across five interest rate scenarios. Models how a bank's P&L shifts as rates move. One command produces a full scenario comparison chart.

Quantitative FinancePythonARIMARisk

github.com/shishaustin

analyst-ai-toolkit

Finance analyst productivity scripts

Standalone scripts that pull live data and do one thing well. Start with the comp table builder — feed it any list of tickers, get a formatted peer comparison table with P/E, P/B, ROE, and margins in seconds.

PythonyfinanceAnalyticsCharting

github.com/shishaustin

ml-model-governance

ML model governance for regulated environments

SR 11-7 and OCC-aligned model card generator plus statistical drift detection (PSI + KS test). Built for data science teams at regulated institutions that need to govern production models without a vendor tool.

Model RiskPythonSR 11-7Drift Detection

All projects use synthetic or publicly available data. Personal research only — not affiliated with any employer. Not investment advice. View all repos on GitHub →