Automated data extraction & verification to accelerate non‑credit program accreditation reviews. Vue UI + TypeScript services orchestrate Retrieval‑Augmented Generation with Neo4j & Pinecone, cross‑checking user submissions against external sources.
Data prep → feature engineering → hyperparameter tuning → evaluation with Gradient Boosting, Random Forest, and Logistic Regression. Included cross‑validation and feature importance to ensure robust generalization.
Processed a decade of data (2.5M rows), engineered features, and delivered Tableau dashboards mapping spatial clusters and seasonal trends to inform policy and resource allocation.
Regression and classification with feature engineering, scaling, encoding; improved accuracy by ~15‑20% via iterative tuning and metrics‑driven validation.
Built a data analytics pipeline in R to transform messy datasets and model outcomes; increased processing efficiency and model accuracy using statistical tests and visualization.
Managed 10k+ customer records and surfaced behavior insights via SQL analytics including basket analysis and basic time‑series techniques.
Automated a recurring billing process (‑90 minutes) with REST API ingestion and Excel transformations to standardize outputs for finance.
Python + JSON plugin to pull queue metrics hourly and publish to dashboards; reduced anomaly triage time during P1 events.
Analyzed migration candidates for a hybrid cloud model and delivered Power BI visualizations after cleansing large server datasets.