OpenAtlas Opportunity Score 92/100
RFA-RM-27-012
NIH PRIMED-AI Data-to-Model Academic-Industrial Partnerships (RFA-RM-27-012)
NIH Common Fund PRIMED-AI opportunity requiring a substantive academic-industrial partnership to develop commercially driven, AI-enabled, image-centered multimodal clinical decision-support tools toward Software as a Medical Device. Earliest submission date is September 19, 2026; applications are due October 19, 2026 at 5:00 PM applicant-local time.
FundingNIH anticipates 6-8 awards. Direct-cost limits are $450,000 per year in the UG3 phase and $800,000 per year in the UH3 phase.
DeadlineEarliest submission date September 19, 2026. Application due October 19, 2026 at 5:00 PM local time of the applicant organization. Late applications are not accepted.
Last verifiedOctober 1, 2026
Required next actionReview RFA-RM-27-012, confirm the required academic-industry partnership structure, required attachments, budget limits, and submission package, then submit by October 19, 2026 at 5:00 PM local time of the applicant organization. NIH states late applications will not be accepted.
Broad Eligibility
D2M-AIP requires a multi-institutional team including at least one academic institution and one for-profit industrial institution. The industry partner must be a substantive co-leader/MPI. NIH states a startup or small for-profit may satisfy the industry role if it has strong potential for SaMD development, evidence generation and commercialization. Academic-prime/company-MPI-via-subaward is acceptable and may be optimal.
Atlas Analysis
Open NIH Common Fund PRIMED-AI Data-to-Model Academic-Industrial Partnerships opportunity for multi-sector teams developing novel image-centered, multimodal AI-enabled clinical decision-support tools. Applications are due October 19, 2026 at 5:00 PM local time of the applicant organization. Strong candidates should combine clinical imaging, multimodal data, AI expertise, and substantive academic-industry participation, with all scientific, budget, and submission requirements validated against RFA-RM-27-012.
The NIH Common Fund is accepting applications for RFA-RM-27-012, PRIMED-AI: Data-to-Model Academic-Industrial Partnerships (D2M-AIP). The program is designed for commercially driven academic-industry teams developing novel AI-enabled, image-centered, multimodal clinical decision-support tools toward Software as a Medical Device.
Timing: the earliest submission date is September 19, 2026. Applications are due October 19, 2026 at 5:00 PM local time of the applicant organization, and late applications are not accepted.
Funding: NIH anticipates approximately six to eight awards. Direct-cost limits are $450,000 per year during the UG3 phase and $800,000 per year during the UH3 phase. The UG3 phase may last up to two years and focuses on tool development and technical validation; the UH3 phase advances development and validation of potential future clinical value.
Partnership requirement: the industry partner must be a true co-leader and serve as a multiple PI with decision-making authority. NIH states this is a hard eligibility requirement. The primary industrial partner must be a for-profit organization of any size with a proven track record or strong potential for medical-software development, regulatory approval and/or commercialization. A startup or small for-profit can qualify, including a structure in which an academic institution is prime and the company principal serves as the industrial MPI through a substantive subaward.
Atlas fit: this is a strong strategic pathway for Dr. Miltie only if N9+ data and platform capabilities can legitimately contribute to an imaging-anchored, multimodal AI/SaMD project. Clinical imaging remains the anchor data type; N9+ cannot substitute for the required imaging-centered AI innovation. A viable pursuit therefore requires an academic AI/imaging partner, a credible clinical problem, access to multimodal data, a commercialization and regulatory pathway, and a defensible role for N9+ as complementary physiologic, remote-exam or connected-health data and/or clinical workflow infrastructure.
No applicant outreach should begin until Atlas/Zoho national deduplication, legal eligibility, open status, prior-contact history and Email Opt-Out checks are complete.
Source verificationOfficial government source · commonfund.nih.govAtlas last verified this record October 1, 2026. Always use the funder's current application materials, amendments and portal instructions as the controlling source.View source and application materials ↗
