Data & Diagnostics Infrastructure

Multimodal clinical data, genomics, imaging, and trial-matching — the highest barrier to entry and the stickiest economics.

Updated May 28, 2026 · 1 companies

Economic Character

Capital-intensive and slower to build, but data scale is the moat. Closed-loop datasets compound with every patient and cannot be scraped or bought. This is where an AI cycle converts into durable cash flows.

This is the layer the core thesis overweights. The moat here survives model commoditization because the asset is the proprietary data loop, not the model on top of it.

Closed-loop genomic, clinical, and imaging data compounds with every patient encounter. It cannot be replicated by a fast follower or a frontier lab, and it strengthens the more it is used. The trade-off is capital intensity and slower ramps — but the retention and defensibility are the highest in the sector.

Companies

Tempus AI high

200+ petabytes of multimodal oncology data, a recently-hit positive adjusted-EBITDA inflection, and positioning as AI infrastructure rather than a diagnostics vendor.

public · ~$7.7–9.0B