AI Clinical Decision Support
AI platforms that assist clinicians and care teams with real-time diagnosis, treatment planning, and patient management at the point of care.
CAPITAL FIGURES ARE MEDIA-EXTRACTED ESTIMATES, NOT VERIFIED FILINGS.
EXTRACTED FROM 25+ PODCASTS & VC NEWSLETTERS · MEDIA-REPORTED FIGURES, NOT VERIFIED FILINGS
Nvidia is becoming the default strategic backer of clinical AI
Nvidia leads the investor table with 50 deals, far ahead of Oak HC/FT (10), Blackstone (8), OpenAI (6) and AMD Ventures (5). Beyond chips, it is moving into healthcare-adjacent equity: it is reportedly planning to invest in French startups including Doctolib, and it is shipping open multimodal agent models such as Nemotron 3.5 Super VL that handle text, image, document and video inputs, the modalities clinical tools need. Infrastructure financiers like Blackstone sit alongside it, and the same Nvidia-plus-private-credit pattern shows up in neocloud financing. The result is that compute, models and capital are increasingly bundled by a single strategic.
Weekly capital swung from $16.8B (week of Aug 10) to $100M (Sep 14) and $312M (Oct 12), while deal counts stayed in a 2-16 range, so a few outsized checks drive the totals. Stage data shows the same split: seed took $5.4B across just 8 deals, while Series A (24 deals, $6.9B) and Series B (19 deals, $2.5B) carry the volume. Most dollars sit in unclassified rounds ($63B), which suggests large, opaque or strategic financings. Scouting signals such as an undisclosed Khosla Ventures round reinforce how much activity is hard to size.
Why it matters · Headline totals overstate breadth, so investors should track median round size and stage mix rather than weekly capital.
Abridge (Epic-integrated, serving Mayo, Duke and Johns Hopkins), Suki, Tandem Health, Heidi Health, Ambience and ClinicFrame all start from scribing and push toward coding, prior authorization and decision support. Adjacent automation such as Ethermed's prior-auth platform (91% first-pass approval) and Brellium and Arintra show the administrative layer being absorbed. The incumbent EHRs, Epic, Cerner and Athena Health, are the integration chokepoints.
Why it matters · Scribe vendors that fail to move up into diagnosis and orders risk commoditization by EHR-native tools.
Nolla Health markets a vertically integrated AI doctor with clinician-in-the-loop oversight, Counsel Health is a full-stack AI-native medical group, and Adaptive Innovations is an AI-native home health provider. Doctronic and NewDays follow the same pattern, owning the care delivery rather than selling software to health systems. This model captures more of the value chain but takes on clinical and regulatory liability.
Why it matters · Full-stack players compete with, rather than sell to, health systems, which changes who the buyer and the competitor are.
Aidoc's $150M Series E is the category's clearest large clinical decision support round, and imaging and diagnostics names such as Cleerly, Rad AI, Bunkerhill Health, CARPL.ai and Butterfly Network point to depth in radiology and cardiology. Oncology tools such as Triomics, Radical Health and Tempus AI follow, and Iambic's CEO explicitly rejects the AI magic wand framing in favor of clinical proof.
Why it matters · Clinical validation and specialty depth, not generic model capability, are becoming the funding criteria.