A production-grade AIOps framework focused on model integrity and autonomous reliability. Features a LightGBM-driven multiclass inference engine via FastMCP, validated against data leakage through NMI analysis. Includes real-time Population Stability Index (PSI) drift monitoring and a closed-loop human-in-the-loop feedback system for sustainable ML

Completed as part of the 365 Data Science Credit Risk Modeling in Python Udemy course. Developed an end-to-end credit risk modeling pipeline for consumer lending, covering data preprocessing, feature engineering, Probability of Default , Loss Given Default , Exposure at Default , scorecard development, model validation, population stability

Classification monitoring

ML model monitoring is not just drift detection. This repo benchmarks PSI, KS, MMD, and ADWIN across real failure scenarios like data drift, label shift, and model decay in production.

An end-to-end credit risk modeling & telemetry pipeline engineered by Srinivasta. Automates underwriting with a traditional WoE scorecard and an XGBoost EWS. Features champion/challenger validation, real-time PSI population stability tracking dashboards, and regulatory compliance audit documentation powered by SHAP Explainable AI.

This project builds a production-grade ML pipeline to classify Near-Earth Objects (NEOs) as hazardous or non-hazardous. It automates data ingestion, preprocessing, model training, monitoring, and drift detection using GitHub Actions, PostgreSQL, MLflow, DAGsHub, and Grafana.

Transform CI/CD failure logs into real-time automated anomaly detection and root-cause classification using supervised machine learning.