On March 15, 2026, we shipped Santi v1 — the first version of SenterME's Signal-to-Strategy engine.
I want to be precise about what that means — and what it does not mean — because the distinction is the product.
Santi v1 is a governed structural diagnostic engine. It is Diagnostic-Ready, operationally hardened, and capable of doing something no other workforce tool in healthcare currently does: tracking structural strain, stabilization, and early recovery in real time, through a continuous, privacy-safe signal layer, at the unit and role level.
That is a narrow description of a specific capability. Here is what it means in practice.
"Santi v1 does something no other workforce tool in healthcare currently does: it tracks the intervention window in real time."
Most detection platforms operate on a binary: something is fine, or something is flagged. Santi tracks four structural states, which reflect the actual arc of how organizational strain forms and resolves.
| State | What it means |
|---|---|
| Stable | Coordination signals within normal range for this unit |
| Watch | Elevated load detected; pattern movement beginning |
| Strained | Strain concentrating; coordination capacity under pressure |
| Critical | High strain; window narrowing; cost of inaction increasing rapidly |
What is absent from this table is equally important: Recovering is not a state Santi declares automatically. Recovery must be earned through signal evidence — because fragile recovery looks identical to stable recovery until the next stressor arrives. Santi monitors recovery as it emerges and flags fragility before it becomes collapse.
The Structural Health Index (SHI) is the primary output of Santi's v1 architecture. It is a deterministic, human-signal-led structural strain score — not a probabilistic estimate. Santi v1 does not make guesses. It produces a governed, versioned output from a defined signal architecture with claim-permission rules built in.
What those claim-permission rules mean in practice:
| Claim type | Governance tier required | Available in v1? |
|---|---|---|
| Unit-level strain state | Reliability gate 1 | Yes |
| Intervention window open/narrowing | Reliability gate 1 | Yes |
| Stabilization detection | Reliability gate 2 | Yes |
| Recovery emerging signal | Reliability gate 2 | Yes |
| Individual exit prediction | Not permitted | Never |
| Causal ROI attribution | Post-pilot evidence required | Not v1 |
| Cross-system benchmarking | Multi-cohort validation required | Not v1 |
This is not a list of limitations. It is a list of governance decisions. Every workforce tool that claims to predict individual exits or guarantee ROI is making claims that the signal quality underlying those tools does not support. We declined to do that. The result is a system that can be trusted — which is a precondition for the participation quality that makes structural detection possible.
The SHI gives a system-level view. The Middle-Load Index (MLI) is the primary diagnostic output of the 90-Day engagement, and it does something more specific.
The MLI localizes where strain concentrates within the middle layer — distinguishing how nurse managers, charge nurses, and frontline supervisors experience load differently, and tracking how that strain behaves over time: stabilizing, cycling, or compounding.
Why this distinction matters: Nurse manager strain and charge nurse strain have different signatures and different timelines. A tool that averages them together will miss the structural load most predictive of cascade failure. The MLI treats these as separate signals with separate thresholds.
By the end of a 90-Day Diagnostic, the MLI produces an evidence-backed, health-system-specific definition of middle-layer load — calibrated to the operational reality of the partner system, not a generic benchmark.
The intervention window is the period between when strain becomes detectable and when cost-effective intervention is no longer possible.
It is not a fixed duration. It is a function of strain velocity, unit conditions, and load trajectory. Santi calculates it dynamically and surfaces it alongside ranked intervention recommendations — ordered by both financial impact and timing urgency.
| Without Santi | With Santi |
|---|---|
| Strain detected at exit interview | Strain detected at signal formation |
| First data surface weeks to months post-departure | Santi is designed to detect strain weeks to months before it surfaces in existing tools |
| Intervention options: reactive, expensive | Intervention options: preventive, cheap |
| Recovery assumed when visible crisis passes | Recovery confirmed through signal evidence |
| Window already closed when leader acts | Window active, tracked, and expiring |
A high-ROI intervention on a closed window is not actionable. That is why Santi ranks interventions by both financial impact and timing — because the order in which you act matters as much as whether you act.
Santi v1 is pilot-ready. It is not category-complete. That distinction is intentional.
The roadmap that follows v1 sequences structural enrichment, recovery breadth logic, and eventually a multi-class sensing architecture incorporating operational and workflow signals alongside the human signal layer. Each class earns its computational role through pilot evidence, not assumption.
We shipped what we could govern. That is not a hedge. That is the right order.
We're actively forming the first cohort for the 90-Day Structural Health Diagnostic. The organizations that move now will help define how this category is validated.