sostriage
AI, clinical logic, and available data
Hybrid artificial intelligence in sostriage combines structured clinical logic with AI-assisted information processing and clearly sourced data. Human professionals retain responsibility for review and the final outcome.
01
A structured approach to information
The quality of an assessment depends on the quality, context, and traceability of the information available. sostriage is designed to organize relevant inputs, identify missing information, and present a structured view for clinical review.
02
Multiple sources, clearly distinguished
Information can originate from the conversation, patient history, supported devices, a helper's observations, or experimental camera-based estimates. Each source is labelled. A camera estimate is not presented as a device measurement, and an unrecorded value is not treated as normal.
03
AI within a defined workflow
The AI layer operates within the product's defined technical and clinical boundaries. It must not silently bypass escalation rules, create unsupported findings, or substitute an automated recommendation for a required clinician decision.
04
Designed for human oversight
The clinician must be able to understand which information was provided, which information was estimated, what remains uncertain, and what action has been confirmed. The goal is to make relevant information easier to organize, review, and act upon. The machine is not the decision-maker.