DEMO MODE — Simulated data from a fictional 500-bed medical center. No real PHI.
Clique
Back to site
TWO SURFACES · ONE PLATFORM

Clique runs hospital AI compute. Pick the surface you live in.

The same infrastructure serves three roles at a hospital: IT keeps it running, clinicians use it to help patients, researchers push the science forward. Each gets a tool that fits their day.

Both demos share the same simulated tenant — St. Mary's Medical Center
Control Plane

Operate. Audit. Govern.

Built for — CIOs · IT · Compliance officers

A real-time dashboard over every GPU, workload, vendor container, enclave, and audit event across the hospital. HIPAA-aligned, continuously verified, and built to stand up to a SOC 2 auditor.

Live GPU / workload / compliance telemetry with 1 s tick
Continuous HIPAA & SOC 2 check engine + one-click PDF reports
Priority lane for emergency workloads (stroke, sepsis, outbreak)
Role-scoped views — Clinician, IT, Researcher
Launch Control Plane demo
~3 min Fully interactive ⌘K · ? shortcuts
Studio

Train. Validate. Deploy.

Built for — Clinicians · Researchers · PIs

Drop a labeled dataset, pick a task, hit Start. The platform auto-configures the architecture to the hardware you have, runs training in a self-spinning container, and hands the validated model to IT with a full audit trail. No code. No config files. No DevOps.

Drag-and-drop DICOM, NIfTI, or CSV — PHI stays on-prem
Auto-configured pipeline (inspired by nnU-Net heuristics)
Self-spinning research containers — zero idle cost
Versioned artifacts, handoff-ready for IT validation
Launch Studio demo
~2 min Guided flow No code
GUIDED SCENARIO · ~90 SECONDS

Simulate a hospital-wide outbreak

Press the button. The Studio auto-trains a triage model on simulated outbreak data, pushes it into shared state, and the Control Plane escalates it onto the emergency lane with full audit trail. Watch how one platform behaves when both sides are live.

Tip: open /control in a second tab before clicking — the new model shows up there live.
Both demos share one underlying platform

Same tenant

St. Mary's Medical Center — one org, one audit log, one set of enclaves.

Same enclaves

Models trained in Studio run in the same attested containers the Control Plane governs.

Same audit log

Every drag-drop, every dispatch, every export — hash-chained, 7-year retention.

Same GPUs

Studio jobs land on the same nodes IT monitors. No duplicate infrastructure.