InsightOS — ML & Causal Analysis Platform
Multipage FastAPI app automating dataset exploration with SHAP explainability, causal inference, and executive summaries to support faster data-driven decisions.
- Python
- FastAPI
- SHAP
- ML
A selection of things I've built.
Multipage FastAPI app automating dataset exploration with SHAP explainability, causal inference, and executive summaries to support faster data-driven decisions.
Multi-label NLP classification on the Kaggle Jigsaw dataset using logistic regression, SVM, Bi-LSTM, and CNN models with exploration of BERT transformers.
Two-part project exploring music data: a popularity predictor using audio features across linear regression, ridge, GAMs, and XGBoost; and a deeper exploration pipeline analyzing trends across genres, artists, and audio characteristics.
Built a Unix shell and multi-stage memory simulator implementing process control, heap allocation, DRAM timing, and LRU cache in C.
R/ShinyApp web app to analyze and display pitching data across 40+ D1 games using Trackman, Rapsodo, and TruMedia.
Automated pipeline using TypeScript to detect, compare, and export forecast Excel files from SharePoint into Tableau-connected sheets.
Finance analysis using Bloomberg Terminal identifying AI opportunities and strategic risks for Capital One in a national case competition.