San Diego, CA dkasat@ucsd.edu LinkedIn Google Scholar deepakshikasat.com
Scientific AI and data lead with experience across predictive modeling, automated scientific workflows, high-throughput screening, and enterprise AI strategy in oligonucleotide drug discovery. Appointed AI Champion for 300+ scientists in a 3,000+ person R&D organization. Built toxicity prediction models that enabled in silico discovery of novel sequences validated in vitro and progressing toward clinical trials. Designed automated hit-picking platforms that reduced processing from days to minutes across multiple therapeutic targets. Bridge between wet-lab biology, machine learning, data infrastructure, and business goals.
AI/ML and Computational Biology: Predictive toxicity modeling (TLR8, ASO safety), automated hit-picking, AI agent development, Python, pandas, NumPy, scikit-learn, XGBoost, matplotlib, seaborn, feature engineering, model validation, reproducible pipelines, FAIR data principles, statistical inference.
AI Strategy and Leadership: AI Champion for 300+ scientists, AI Task Force leadership, enterprise AI adoption, use case packaging and scaling, responsible AI governance, scientist training and enablement, cross-functional alignment.
Automation and High-Throughput Screening: Automated liquid handling (Tecan Fluent, Hamilton STAR, Echo, Integra), multi-liability hit-picking, plate-based screening, dose-response workflows, operator guides, troubleshooting systems, Benchling and Labguru.
Assay Development and Biology: High-throughput NGS and cell-based assays, flow cytometry, ELISA, western blotting, immunoassays, RT-qPCR, RNA accessibility, uptake, target engagement, hepatotoxicity screening, complement activation, primary neurons, cardiomyocytes, iPSC-derived models, confocal and high-content imaging, CRISPR workflows.