Private AI

Private AI, operating under the product name Limina, is a data de-identification company that develops context-aware PII identification, redaction, and anonymization software for enterprise organizations. Unlike pattern-matching tools, Limina reads context to catch names that do not look like names, entities across sentences, and credit cards spoken across conversation turns. The platform detects over 50 entity types including PII, PHI, and PCI across 52 languages, handling messy data including ASR errors, OCR mistakes, handwritten forms, and conversational disfluencies. Limina supports multiple transformation modes including redaction, pseudonymization, synthetic PII generation, and generalization, producing expert determination-ready outputs for HIPAA, GDPR, CPRA, and other global privacy regulations. The container-based solution deploys in the customer's VPC or on-premises, ensuring data never leaves their infrastructure. The platform processes billions of API calls per month in production, achieving 99.5% accuracy on physician conversations and reducing medical inquiry response time from 48 hours to minutes for Boehringer Ingelheim. Founded in 2019 and headquartered in Toronto, Canada, the company has raised approximately $11.3 million in Series A funding from investors including M12, Shasta Ventures, Y Combinator, Differential Ventures, Hyperplane, and BDC, with a team of 149 employees.