AI Innovation Institute Technology Solutions

The University of Connecticut is soliciting proposals through a solutions-based Request for Proposal (RFP #CDR010826) to design, build, supply, install, support, and continuously enhance state-of-the-art AI technology resources and managed services for the establishment of an AI Innovation Institute. The procurement seeks a qualified partner to deliver comprehensive AI computing infrastructure including approximately 400 latest-generation GPUs and 400 AI accelerator nodes with minimum computing performance exceeding 30 petaFLOPs, approximately 15 petabytes of storage with anticipated annual growth of 5 petabytes, and high-speed redundant data fabric supporting within-cluster networking. The solution must support three logically segregated enclaves—a General Research Enclave, a HIPAA-compliant enclave for Protected Health Information, and a Federal Compliance Enclave meeting NIST SP 800-171 and FISMA requirements at the high level—across test, UAT, and production environments. The awarded proposer must provide 24/7/365 managed service support with 4-hour onsite response time for physical equipment issues and 4-hour response time for cloud service incidents, along with a three-year warranty extendable for a minimum of two additional years. Proposers have discretion in proposing on-premises, cloud-based, or hybrid deployment modalities; however, the University has not identified data center space for on-premises infrastructure and will not provide direct-to-chip liquid cooling support at its facilities. Evaluation criteria are weighted as follows: Technical Capability and Demonstrated Success (25 percent), System Architecture, Project Plan, and Timeline (25 percent), Financial Proposal including pricing and long-term cost-effectiveness (30 percent), and Value-Added Considerations such as AI Ecosystem Development, Joint Research Programs, and AI Education and Training (20 percent). The pre-proposal conference was held on January 21, 2026 at 2:30 PM ET, with questions due by February 10, 2026 at 2:00 PM ET and proposals due February 18, 2026 at 2:00 PM ET; the initial contract term is four years with five optional one-year extensions available, potentially extending the contract to nine years.

The University encourages participation from Small Business Enterprises (SBE), Minority-Owned Businesses (MBE), Woman-Owned Businesses (WBE), and Disability-Owned Businesses (DisBE) certified by the Connecticut Department of Administrative Services. All proposers must comply with mandatory submittal requirements including a Campaign Contribution Certification form, Bidder Contract Compliance Monitoring Report, acknowledgment of the University's Vendor Code of Conduct, Connecticut Economic Impact Form (DAS-46), and a Non-Collusion Affidavit, and must provide three references from comparable organizations with preference for higher education institutions demonstrating technical capability in designing, deploying, and managing AI systems at comparable scale and complexity. Proposers must integrate with the University's existing identity management systems, specifically Microsoft Entra for general-purpose and HIPAA enclaves and Microsoft Entra Government tenant for the federal compliance enclave, and may leverage the University's existing Microsoft Azure agreement, including a GCC High subscription, to support the federal enclave. The awarded proposer will be required to sign a Business Associate Agreement (BAA) for HIPAA and federal enclaves and must address compliance artifacts including System Security Plans, Plans of Action and Milestones, encryption and key management approaches, and post-quantum cryptography considerations. The University has not disclosed a specific budget or funding source but reserves the right to make either a single award or multiple awards as deemed in its best interest, with pricing proposals requiring itemized breakdowns for one-time costs, recurring operating costs, per-user or usage-based charges, and compliance-specific expenses for each enclave.