Robotics Startup Remy AI Targets Under-Automated Warehouses
YC-backed startup Remy AI is developing dexterous mobile robots for 3PLs and small to medium-sized businesses. The company claims its robots can adapt to variable warehouse tasks like picking and packing without retraining and at a 50% lower cost than existing solutions. The startup aims to serve the 80% of U.S. warehouses that are reportedly under-automated.
- The founding team consists of CEO Oscar Brisset, who previously advised Fortune 500 logistics companies at BCG, and CTO Ben Kaye, who holds a PhD in Machine Learning from Oxford with a focus on 3D computer vision. - Remy AI is part of Y Combinator's Winter 2026 batch and was founded in 2025 in San Francisco. - The company's technical approach centers on a proprietary machine learning model and a custom training pipeline designed to learn new, high-dexterity tasks with less data than standard models require. - The target market of under-automated SMB and 3PL warehouses often faces high implementation costs for automation, with comprehensive solutions typically ranging from $15,000 to $50,000 and enterprise systems exceeding $100,000. - The concept of "agentic AI," which Remy AI's adaptable robots align with, is a significant trend where systems move beyond executing programmed tasks to autonomously planning and acting to meet goals, with Gartner naming it a top technology trend for 2025. - Key challenges for 3PLs, a target customer, include slow order fulfillment from manual processing, inventory discrepancies, and a need for scalable, flexible automation that can adapt to diverse client needs and seasonal demand shifts. - Competitors in the broader warehouse robotics space include Locus Robotics, which focuses on autonomous mobile robots (AMRs) that assist human pickers, and GreyOrange, which provides AI-driven software to orchestrate robots and human workers. - The robots are specifically designed to be slotted into existing workstations, targeting "brownfield" automation—upgrading existing facilities rather than requiring the construction of new, purpose-built automated warehouses.
