Deep Viral Mehta

About Deep Viral Mehta

With a strong foundation in machine learning, deep learning, and data analysis, I am an…

Experience

1. Data Scientist

   Prince William Water

   Jun 2024 — Present · VA, US

   Water Consumption Prediction Model: Developed a predictive model to forecast future water usage needs with an error margin of ±3.5%, enhancing procurement planning. Integrated additional factors, including holidays and humidity, to improve seasonal accuracy by 15%, allowing predictions up to one year. Achieve an R-squared of 0.89 using advanced machine learning models (Gradient Boosting, XGBoost, LightGBM) and neural networks (PyTorch). • Methanol and Ferric Chloride Usage Prediction: Working on models for methanol and ferric chloride demand in wastewater treatment, factoring in peak usage patterns to optimize ingredient planning. Plan to reduce chemical usage by 20%.• Data Processing and Visualization: Designed a data pipeline that decreased data processing times by 40%, allowing faster insights and more agile decision-making. Leveraged PowerBI to produce visualizations, increasing report readability as per stakeholder feedback.• Dashboard and Reporting Automation: Centralized data and automated dashboards improved workflow efficiency by saving man-hours, and providing critical real-time insights.• System Integration and Data Sharing: Integrated disparate business systems to reduce data silos, increasing data accessibility across departments, and fostering improved collaboration and data-driven initiatives.• Master Data Management and Asset Analytics: Helped curate a master data repository that underpinned enterprise-level reporting, yielding 20% higher data consistency.