top of page

Cloud Cost Anomaly Detection & Forecasting

  • Handled 1M+ real-world scale records: Built and cleaned a synthetic cloud usage dataset with 12 months of logs, including missing values, duplicates, outliers, and inconsistent formats to simulate real IT data challenges.

  • End-to-end data science pipeline: Applied anomaly detection (Isolation Forest) and forecasting (Prophet) in Python, uncovering spend anomalies and predicting cloud costs with seasonality and trend insights.

  • Business-ready Power BI dashboard: Delivered an interactive dashboard for IT managers to track anomalies by service/team, forecast spend, and make proactive cost optimization decisions.

Python logopng_edited.png
power bi_edited.png
bottom of page