San Francisco, California, United States
•Coded a Value-at-Risk calculator engine in Python for all commodities that sped up output by saving time; created web application in Streamlit and hosted it on Azure; enhanced tools for the stress-testing of energy portfolio using python. •Devised end-to-end PowerBI dashboard for power curve validation; implemented ETL processes using Snowflake for real time dashboard updation; coded SQL queries to reduce time taken for data extraction. •Engineered ad-hoc tools for quantitative traders for desk-specific functions; improvised decision-making for desk by as measured by portfolio managers; regulated analytical models for market risk team to calibrate risk limits.
• Programmed custom predictive model to improve sales by 15% during covid; identified sales trends using time-series analysis and streamlined supply chain operations to implement model solutions; trained 2 teams to maintain model and incorporate anomaly. • Analysed 50,000+ spare parts of biomass boilers using Tableau dashboard to identify supply chain bottlenecks; reduced inspection costs and time by 12% for the operations department by analysing 1MM+ purchase orders to fortify supplier and vendor network. • Improved predictive power of existing analytical models to forecast defects in purchase orders by at least 35%; devised web application in Streamlit and hosted on Docker to increase efficiency between supply chain and operations departments.
• Constructed a Decision Tree Regressor to forecast internal energy consumption to achieve an RMS error of 3.91 as compared to a baseline Linear Regression model; hyper-tuned the model to minimize RMS to 0.225 and increase the r-squared score to 0.95. • Developed visualization tools to analyse 150+ crash reports to extract key metrics by feature engineering for projected energy absorption during crash events; increased performance of analytics team through data engineering. • Redesigned automation tools to reduce computational costs by a margin of 20% across cross-functioning teams; led a 5-member intern team to successfully develop case study on Tesla crash pattern analysis
Led a team of 20+ members to enhance the contribution of engineering students towards Industry 4.0 by partnering with a higher executive from Mercedes Benz. Executed an unprecedented competition which saw participation of 400+ students with their projects in Mechatronics. Jointly increased the annual output hosted, by 20%.