Harini Vasan

Reduced manual validation effort by 35% by building automated Python preprocessing and feature-engineering pipelines for AI/ML datasets. Accelerated model regression detection by engineering precision, recall, and F1 evaluation dashboards for real-time performance monitoring. Improved prediction reliability by implementing live model-drift monitoring and automated retraining workflows.

AI Engineering Intern - Crescent Innovation & Incubation Council - Chennai, TN

(2024-05 - 2024-06)

• Reduced manual validation effort by 35% by building automated Python preprocessing and feature-engineering pipelines for AI/ML datasets.
• Accelerated model regression detection by engineering precision, recall, and F1 evaluation dashboards for real-time performance monitoring.
• Improved prediction reliability by implementing live model-drift monitoring and automated retraining workflows.

Data Science Intern - Infinity Cosmos Technology - Remote

(2024-06 - 2024-07)

• Influenced quarterly planning decisions by analyzing 100K+ records to uncover revenue and churn trends through EDA and statistical modeling.
• Reduced manual reporting effort by 70% by automating KPI dashboards in Power BI using DAX and Power Query workflows.
• Improved ML pipeline throughput by 40% by engineering SQL preprocessing workflows with CTEs and window functions.

B.Tech - Artificial Intelligence & Data Science - Adhi College of Engineering and Technology (2022 - 2026)