The request will use the currently selected route and adapter settings.
Showcase home
Interactive workspace
Cloud-connectable
Product-facing
Operational workspace
Run CKD risk prediction without leaving the product surface.
This interface is designed as the execution layer of the platform: structured intake on the left, returned prediction evidence on the right, and adapter controls that let the same frontend move from mock mode to live backend and cloud routes.
Mock mode
Prediction flow
Move through one structured operating sequence
Step 1 / 4
Mode selection
Choose how the request should enter the platform
Active mode
Clinical Intake Mode
Use clinically familiar variables for a product-facing workflow. This mode assumes the backend accepts native clinical inputs or performs intake-to-model translation.
Serving route
/predict/clinical
Configured in the adapter settings panel below.
Input scope
14 fields
The active mode controls which schema is collected and submitted.
Case context
Frame the prediction request before you send it
Input schema
Clinical intake form
This mode is designed for a product-facing workflow and can later connect to a backend that performs clinical feature translation or direct model inference.
Review and run
Check the request before sending it to the prediction route
The review step helps confirm completeness before you run the request.
No request generated yet.
Prediction output
Returned score and serving result
Risk score
--
No prediction yet
Prediction label
--
Model version
--
Timestamp
--
Serving route
--
Capture
Intake collected
Route
Adapter dispatch
Response
Prediction returned
Submit a request to generate a structured interpretation summary for the active mode.
The workspace will propose an immediate action cue after the prediction result arrives.
A concise report snapshot will be assembled from the latest response, route, and signal summary.
Decision support
Clinical guidance, top signals, and report snapshot
Clinical guidance
Submit a prediction to generate clinical notes.
Top signals
Top contributing features will appear here after prediction.
Report snapshot
Case
Pending
Mode
Pending
Summary
No prediction summary available yet.
No explanation items yet. Run a prediction to build the contribution path.
Adapter settings
Backend and cloud connection
The same static frontend can hand requests to a local backend, API Gateway, Lambda, or another cloud-served prediction route once the adapter is pointed at a live endpoint.
Export report
Package the current result into a shareable prediction report
Run a prediction to assemble a structured report with summary, guidance, signal explanation, and serving metadata.
No report preview yet. Submit a prediction to build the report surface.
No report preview yet. Submit a prediction to build the report surface.