Operational Intelligence for Patient Care.
Florence uses KEIS to remember the patient as a living clinical environment: vitals, activity, sleep, medications, surroundings, care actions, and outcomes. The result is a clear daily read for the care team: what has changed, why it matters, and what should be watched next.
KEIS © is an operational intelligence engine developed by WiCis. In Florence, KEIS is not presented as only an evironment system. It is the shared architecture underneath the medical product: a memory-based engine that compares today’s patient state with prior patterns, prior interventions, and prior outcomes.
From signals to care priorities
The interface stays simple. The intelligence layer underneath remains recursive: observe, remember, compare, prioritize, and learn from every care episode.
Sense
Continuously collects vitals, symptoms, activity, sleep, medication events, and environmental context.
Remember
Builds patient-specific memory from trends, episodes, care actions, and recovery or deterioration patterns.
Compare
Finds similar past situations for this patient and across relevant cohorts when clinically appropriate.
Prioritize
Separates noise from operationally important change, showing what requires attention now.
Learn
Care-team actions and outcomes flow back into memory, improving the next operational read.
The Patient is the Environment.
In the original KEIS architecture, the system interprets current conditions through environmental memory, operational memory, archive intelligence, and curator input. In Florence, those same concepts become patient-centered.
Current conditions are the patient’s live vitals and observations. Environmental context becomes room, sleep, mobility, meals, hydration, weather, social factors, and care setting. Operational memory becomes what happened after prior interventions.
- Not a generic dashboard, but a memory-aware clinical intelligence layer.
- Not an alarm system, but a prioritization system for care teams.
- Not prediction alone, but explanation linked to prior patient episodes.
- Not black-box AI, but clinical judgment supported by traceable memory.
- Every observation and outcome improves the next read.