London – – September 30, 2026 -- Everpure (NYSE: P) has launched new platform capabilities that cut large language model inference latency by up to 20x and give enterprises native access to governed data for AI agents, the storage and data management company announced. The updates, rolling out in October, extend the company's Data Primacy architecture first introduced at Pure//Accelerate in June.