An EHR built on a graph care model — every patient's case is a longitudinal knowledge graph, compressed into a one-page brief for the care team.

DocDoc is an EHR built on a graph care model. Instead of a fixed schema of tables and forms, each patient's case is modeled as a longitudinal knowledge graph: conditions, symptoms, hypotheses, tests, observations, actions, and outcomes are all nodes, and the relationships between them — supports, refutes, contributes to, measured by, precedes — are typed, time-bounded edges.
The data model is deliberately general: one polymorphic node table and one edge table, with typed JSON properties. It's inspired by the entity–relationship generality idea from Google's network-topology modeling — describe everything as entities and relationships, and new clinical concepts become new types rather than schema migrations.
On top of the graph, a Brief Studio renders a calm one-page Doctor Brief: today's goals and calls to action, progressive reveal from summary to evidence, and the causal map of how symptoms, factors, hypotheses, and actions connect. An admin workbench edits the underlying graph directly.