Animesh Jain

Designed and deployed end-to-end time-series analytics pipelines for large-scale sensor data (TDP, PMW, vibration, temperature, power), enabling high-quality insights across multiple pumping stations. Achieved 90-95% forecasting accuracy for power, vibration, temperature, and flow parameters by integrating Seasonal-Trend Decomposition (STL)-extracted trends with Prophet, ARIMA, and SARIMA time-series models. Developed hybrid forecasting pipelines combining trend decomposition + Prophet/SARIMA, enabling highly reliable fault prediction and early anomaly detection for pumping systems. Built advanced outlier detection modules using MAD, rolling statistics, and comparative signal validation, significantly improving data trustworthiness for forecasting models. Delivered high-quality analytical dashboards visualizing trends, seasonality, forecast intervals, anomaly bands, and fault-risk zones, enabling operations teams to make data-driven decisions.