Verified public evidence
CerviGuard live surface, product screenshots, public repository, model/profile hub, and product route.
Evidence baseline for product review
This page provides the current public evidence set: product timeline, pilot context, and KPI framework. Numeric KPI publication remains gated until cohort definitions and reporting windows are finalized.
Document status: Current public baselineLast updated: 2026-02-17Owner: Product + Commercial + Data
Reported outcomes are cohort-specific, method-bounded, and include explicit limitations. KPI percentages will not be published without denominator and date-window context.
Evidence status
Qualified public evidence
Draft regulatory material, security baseline pages, and methodology notes that require dated context.
Evidence gaps
Clinical outcome percentages, commercial traction metrics, and final regulatory identifiers are not published yet.
Pending metrics
KPI values remain withheld until cohort definitions, denominator rules, and reporting windows are finalized.
Public milestone timeline
2019-2020
First CerviGuard pilot phase
Initial downloadable Win32 TensorFlow-based pilot workflow documented in MDR draft references and planning artifacts.
2022
Research and project track consolidation
Multiple projects and research tracks were finalized before broader platform expansion.
2024
Company launch and platform preparation
Company launch milestones and online product transition documented in internal/public planning artifacts.
2025
Public product evidence expanded
SmartClover expanded public product evidence through CerviGuard screenshots, product routes, and supporting trust pages.
Pilot context and methodology notes
- Pilot evidence is currently published through milestone and methodology summaries.
- Outcome publication requires fixed cohort windows and approved numerator/denominator definitions.
- Clinicians review AI outputs before confirming follow-up actions.
KPI framework (values pending publication)
| KPI name | Numerator | Denominator | Cohort window | Method | Caveat |
|---|---|---|---|---|---|
| Case review turnaround improvement | Cases completed inside target review window post-deployment | All eligible cases in selected cohort | Pending publication | Pre/post cohort comparison with fixed inclusion criteria | Workflow mix can vary by site |
| Follow-up adherence uplift | Cases with completed follow-up within defined interval | Cases assigned follow-up action | Pending publication | Operational follow-up completion tracking | External care-path variability applies |
| AI-assisted triage agreement rate | Cases where AI support category aligns with clinician final category | Cases containing both AI output and clinician adjudication | Pending publication | Blinded adjudication and reconciliation rules | Not a standalone diagnostic-accuracy claim |