Jason Natale

WORK HISTORY

EDUCATION

ABOUT JASON NATALE

Jason is a Humanist who believes that the Singularity is nearer than we think and that positive futures must be thoughtfully created.He has 7 YoE building models and solutions across the entire ML lifecycle. As a results-driven engineer, Jason has a long track record of identifying and deeply understanding problems, delivering value, and iterating rapidly on lean, fast-moving teams operating in challenging domains. He has spearheaded scientific contributions of his own from ideation to deployment, while also supporting and mentoring ML Scientists with data & platform engineering along the way.Most notably, Jason trained, deployed, and presented a Multi-Agent Reinforcement Learned (MARL) model for autonomous stem cell colony control. The MARL agents, trained in a parallelized In Silico simulation which Jason also built, were successfully deployed Sim2Real to Cellino\'s biomanufacturing system and doubled production capacity estimates.Additional career highlights also include the training of a ViT decoder from SAM2 Foundation Model embeddings to regressively model cell culture densities, co-inventing a variety of morphological management techniques for stem cell colonies, prototyping a Few-Shot In-Context Learning (ICL) Pluripotency classifier with the GPT-4o Vision Language Model (VLM), designing and building a graph-based cell culture tracking and visualization system, scaling RL systems for billions of products at Wayfair, and founding an Augmented Reality Metaverse startup.Jason is currently seeking another opportunity to build autonomy zero to one on an ambitious team with an incredible vision and strong technical leadership.