15 GHP Q4 2024 Oct24434 ccelerating the development and deployment of artificial intelligence (AI) and machine learning (ML) solutions to achieve real world impact is key to competitiveness across the healthcare, biopharmaceutical, and life science sectors. AI/ ML solutions hold great promise to improve patient finding, optimize patient outcomes, improve clinical trials efficiency and recruitment, achieve operational efficiencies, and enhance financial performance. To achieve these goals, they are leveraging EscrowAI®, the #1 most innovative healthcare AI platform developed by BeeKeeperAI®. EscrowAI addresses critical challenges for both the data stewards and algorithm developers, by enabling models to securely compute on protected, real-world data. The platform ensures regulatory compliance while protecting patient privacy and AI model intellectual property (IP). As a result, purchasing, security, compliance approvals and contracting are streamlined by as much as 80%, making EscrowAI an invaluable tool for organizations looking to leverage sensitive data to achieve real world impact more quickly. EscrowAI is a development, validation, and deployment monitoring platform that enables data stewards (i.e., healthcare delivery organizations) and algorithm developers (i.e., pharmaceutical and life science companies) to securely collaborate while maintaining protections for data sovereignty, individual privacy, and AI/ML model IP. EscrowAI integrates and automates privacy-enhancing technologies (PET) including Trusted Execution Environments (TEE) with confidential computing and end-to-end encryption for data and IP within a collaboration workflow. The platform has been granted seven U.S. Patents, with an additional 20 patent applications pending. As a result of BeeKeeperAI’s unique approach and impact, EscrowAI has been named the 2024 “Most Innovative Healthcare AI Platform” in the United States by Global Health & Pharma Magazine. • Real-time ML Lifecycle Collaboration Workflow & Tools: Delivers an integrated user experience and tools for data stewards and algorithm developers to perform training, validation, and monitoring at a single site or on a federated basis. • Protecting Individual Privacy More Securely WITHOUT Deidentification: With EscrowAI, AI/ML models are brought to the data steward’s environment to confidentially compute on the encrypted, personally identifiable and protected health information (PII and PHI), resulting in improved algorithm precision and real world performance in less time. The risk of re-identification or personal information and data getting into the wrong hands is eliminated with EscrowAI. • Data Type Agnostic: Any data type may be computed upon within the TEE including imaging, structured, unstructured, video, genomics, waveforms, etc. • End-to-end Encryption: Data and IP are protected with robust encryption methods (AES-256) at rest, in transit, and in use within the TEE. • Supports GPU for large complex data and AI models: Depending upon the size of the algorithm and the data set, the algorithm developer selects the optimal confidential CPU BeeKeeperAI Is Named Most Innovative Healthcare AI Platform of 2024 A and GPU virtual machine computing environment. Only pay for the compute power you need. • Security Attestations within the TEE: Only predefined, verified outputs (i.e., reports or models) are released from the TEE; outputs are checked for PHI exfiltration. • Immutable Computing Record: The data and algorithm version are tracked and reported for each computing run and stored in an immutable record of the algorithm’s development process. This record can support assessments for model drift and can also serve as a security artifact to detect any third-party tampering with the algorithm or data. • Regulatory Compliant: Provides protections for data sovereignty and individual privacy for the US, Canda, and EU. “We are tremendously honored by this recognition. Since we spun-out from UCSF, our mission at BeeKeeperAI has been to improve health and healthcare worldwide by accelerating the development and deployment of trustworthy, reliable, responsible AI. We recognized that secure access to protected, real-world data would be key to that goal which led to the development of our EscrowAI platform. Leveraging the latest privacy-enhancing technologies, we have built and deployed a platform that dramatically reduces the time and cost for a pharmaceutical or life sciences company to securely compute on sensitive patient data (including PII and PHI) while also protecting the IP in their model. Companies can now develop, validate, and deploy AI in a few months, not years,” said Dr. Michael Blum, MD, Co-founder and Chief Executive Officer at BeeKeeperAI. “The current default method for protecting patient privacy is data de-identification. While it can be used for early-stage AI development, de-identification doesn’t allow an AI developer to demonstrate real world performance, nor does it truly protect sensitive patient data. The combination of our privacy-enhancing capabilities has advanced EscrowAI beyond anything else available in the marketplace today. Leveraging TEEs with confidential computing and end-to-end encryption ensures data privacy, sovereignty, and intellectual property protection and dramatically accelerates time to market for AI/ML solutions. No other approach – not even federated learning – can come close to the protections and data type and model flexibility EscrowAI provides without compromising usability and security,” Dr. Blum added. EscrowAI is available in the Azure Marketplace or directly from BeeKeeperAI. Learn more at BeeKeeperAI.com or follow us on LinkedIn Contact: Mary Beth Chalk Company: BeeKeeperAI Website: https://www.beekeeperai.com/
RkJQdWJsaXNoZXIy NTY1MjM3