Parametric builds robots that use reinforcement learning (RL) to automate repetitive physical work with 3x higher task reliability than baseline SOTA models. We're folding laundry today for one of the largest wash-and-fold operators in San Francisco. To do this, we built an RLHF pipeline for robotics similar to those used by frontier language labs. The team led autonomy at Parallel Systems (a $100M+ raised Series B deep tech), and scaled ROI at Facebook for its recommendations.