Neha Annamalai

Graduate Research Intern

Deutsche Bank

Nov 2023 - Current (2 years 10 months)

• Optimized asset allocation strategies using advanced reinforcement learning techniques, such as Deep Q Learning and Policy Gradient, with rigorous hyperparameter tuning, leading to a 13% increase in ROI. • Engineered customized reward functions using advanced Temporal Difference Learning algorithms to dynamically adjust portfolio risk, resulting in a 22% reduction in risk exposure and bolstering overall risk management precision. • Integrated Proximal Policy Optimization and thorough market analyses of equities, bonds, and commodities to inform trading decisions and boost portfolio efficiency by 18%.

Data Manufacturing and Analyst

Onto Innovation

Nov 2022 - Aug 2023 (10 months)

• Developed a manufacturing execution system application, facilitating seamless data communication between cleanroom factory and production management, resulting in 15% increase of production efficiency. • Streamlined data sources and collection methods, cutting down product cycle time by 20% and improving tool build sequence by 35%, leading to faster market delivery. • Optimized parallel processing techniques and utilized node distribution in computing frameworks, including Hadoop and Spark, resulting in a significant 30% reduction of processing time for large data sets. • Presented 25+ interactive dashboards using Tableau, PowerBI & Seaborn to communicate idle time, supply chain deficiency, escalation, and material request shortage, l

F

Data Science Intern

Fanplayr, Inc.

Jun 2021 - Sep 2021 (4 months)

• Executed behavioral customer segmentation by identifying clusters of high converting users with a sizable unconverted population. Determined appropriate targets for optimizing conversion and product positioning. • Deployed Google Cloud AI for model training and strategic feature engineering, improving correlation importance scores by 25% using recursive feature elimination, which streamlined model performance by 18%. • Built 100+ BQML K-Means clusters and visualization models for customer profiling. Assessed trends in user activity by identifying populous segments. Predicted ideal cluster model ratio of 3:7 for converted to total users. • Conducted rigorous statistical analysis to validate effectiveness of A/B testing strategies, incr

Undergraduate Researcher

Santa Clara University

Jan 2021 - Jan 2023 (2 years 1 month)

• Modified the machine learning lifecycle with graph neural networks to transform node representation into lower dimensional spaces through implementing local linear embedding and t-SNE, achieving feature space reduction. • Pioneered innovative method to replace meshes with point clouds for classification, leveraging TensorFlow for efficient distance correlation; increased classification accuracy by 25%, reduced processing time by 40%. • Successfully achieved a notable 30% reduction in feature space, contributing to increased computational efficiency and scalability in machine learning applications.

About

Detail

Timeline

Résumé