DEEKSHA VISHNOI

• Researched and implemented Performance-conditioned Diverse GANs (PcDGANs) for solving inverse design problems using the Cahn–Hilliard mechanical dataset. • Explored continuous conditional generative models, including GANs and diffusion-based architectures, for property-conditioned microstructure generation. • Built and optimized PyTorch and TensorFlow training pipelines, improving runtime efficiency and model stability. • Evaluated model performance using label error, diversity, and likelihood metrics.

• Automated data workflows and reporting pipelines using Python, SQL, Jethro, and Tableau improving data delivery speed and accuracy across departments. • Conducted performance testing and dashboard optimization, reducing data latency by 70%. • Collaborated with Product Owners, Department Managers, and VPs on proof-of-concept projects, increasing process efficiency by 20%. • Created 60+ custom reports in under three quarters, trained team members on BI tools, and contributed to analytics best practices.

• Developed advanced clinical Tableau dashboards for oncologists, anesthesiologists, and researchers at a Cancer Research Centre, enabling data-driven insights into cancer treatment outcomes and patient care. • Performed advanced analytics, including Progression-Free Survival analysis, TabPy-based statistical testing, and Dose Volume Histograms, supporting precision treatment decisions. • Built complex dashboards using data blending, parameters, and custom geocoding for regional treatment analytics.

• Designed, developed, tested, and delivered end-to-end BI solutions for Sales and Pharma analytics projects within the Life Sciences domain, adhering to industry best practices. • Administered Tableau environments (Production, Test, Development), ensuring stable operations and minimal downtime. • Orchestrated ETL workflows using Talend (TAC, Open Studio) with parameterized scheduling for automated data pipelines. • Deployed large-scale Tableau Server dashboards on Amazon Redshift.