Pinned Loading

1. home-credit-scorecard home-credit-scorecard Public

   End-to-end credit scorecard — WOE/IV feature selection, logistic regression vs XGBoost champion/challenger, ECOA adverse action codes, FastAPI scoring endpoint, DuckDB + dbt SQL layer. Gini 53%, KS…

   Jupyter Notebook

2. doc-intelligence-pipeline doc-intelligence-pipeline Public

   RAG pipeline for unstructured document intelligence: layout-aware parsing → ChromaDB vector store → dbt medallion architecture → REST API. 87% F1, <30s/doc, live on Streamlit.

   Python 1

3. Enterprise-Predictive-Maintenance-IoT-Sensor-Optimization- Enterprise-Predictive-Maintenance-IoT-Sensor-Optimization- Public

   Built an end-to-end MLOps pipeline (Databricks/PySpark → Snowflake medallion architecture) for turbofan engine failure prediction using NASA CMAPSS sensor data — starting from structured EDA and hy…

   Jupyter Notebook 1

4. DoorDash-Merchant-Intelligence-Platform DoorDash-Merchant-Intelligence-Platform Public

   Production-grade lakehouse analytics platform: PySpark → Delta Lake → Trino → dbt → Great Expectations → Airflow → Sigma. Built to mirror DoorDash's production data stack.

   Python

5. credit-risk-analytics-pipeline credit-risk-analytics-pipeline Public

   Multi-source data pipeline testing whether interest rates predict credit stress better than unemployment (r=0.80 vs r=-0.28). Built with Airflow, Python, ANOVA, and Tableau.

   HTML 1

6. federal-contract-rag-pipeline federal-contract-rag-pipeline Public

   Dual-method federal contract cost estimator: parametric (CER) + analogy estimating with automated disagreement detection. Validated live against USASpending.gov API. 29/29 tests passing.

   Python