ARNAV AGARWAL

Project – Predictive Modeling

• Increased accurate athlete race predictions by 15% using machine learning-based performance indexing model with statistical techniques and Python in 3 months.
• Spearheaded cross-functional team of 5 -10 members to train, validate, implement model, and provide feedback to end users.
• Optimized machine learning models' recall and precision by ~13% using statistical analysis techniques.

Project – Data Warehousing

• Implemented and maintained centralized SQL database with 94.5% data transfer success rate across departments.
• Migrated data from on-premise warehouse to AWS Redshift, resulting in 15% reduction in infrastructure costs.
• Maintained database uptime of 96%, fostering data integrity, scalability, and remediation of performance issues; implemented updates with 97% success rate.

Project – Algorithm Optimization

• Resolved limitations in index methodologies, increasing ML model accuracy by 5%, using data pre-processing and regression analysis.
• Optimized predictive outcome of existing indexes by ~12% using back testing of algorithms and hyper parameter tuning.

Project – Customer Analytics

• Automated sales pipeline visualization, reducing manual work by 68% using SQL and data visualization on 2M+ rows.
• Boosted monthly revenue by 30% QoQ through partnership with CCO to develop data-driven go-to-market strategies.

Project – Marketing Analytics

• Identified $2M in sales opportunities for 60+ sales reps by creating and launching whitespace and greenfield model.
• Developed and sustained data-driven visualizations using Tableau, providing business partners with average of 4-5 insights per visualization on key metrics and performance indicators.

• Wrote production-ready code with fluency in modern front-end and back-end frameworks.
• Assisted with creating cross-platform UI web components and built 13+ responsive applications using front-end frameworks.
• Used software tools in cooperation with continuous integration and test automation and supported software reuse and refactoring.