Built a predictive inventory system for SAIL, training 50 material-specific Prophet models on 73K+ daily records to forecast demand, predict stockout risk, and generate lead-time-aware reorder recommendations.

An autonomous data intelligence platform that transforms raw datasets into actionable insights through AI-assisted analysis, forecasting, anomaly detection, root-cause diagnosis, and automated executive reporting.

Built a customer retention intelligence platform using XGBoost and SHAP to predict churn, explain account-level risk, segment customer behavior, and simulate pricing strategies for revenue impact.

Built a real-time computer vision attendance system combining DNN face detection with LBPH recognition, low-latency camera processing, automated identity logging, and SQLite-based attendance management.

Built an autonomous finance controller achieving 96.2% automated reconciliation across multi-source transactions, with 100% precision and 14ms processing latency, plus ERP ledger sync and treasury forecasting.