Research
Open methods, private projects
Evidence foundation
Alive's defaults, rubric and citation checks rest on a documented evidence foundation: a registry of literature sources and their terms of use, a citation-verification design grounded in measured rates of AI citation errors, an evaluation rubric mapped to established digital-health and implementation-science frameworks, and cited parameter ranges for each reference world. Every citation in it was opened and checked; items that could not be verified are listed as such.
Research sharing is opt-in
Sharing is off by default. You can opt in at two levels: approved aggregate metrics and high-level metadata, or additional anonymized structured traces. You can revoke at any time; revocation stops future use, though aggregates already published cannot be withdrawn. Alive does not train models on your submissions.
What we intend to study
- Which implementation failures recur across kinds of health innovation.
- How well automated findings agree with expert reviewers.
- How simulation results compare with outcomes later observed in real deployments.
Any study involving human subjects or real clinical data will go through ethics review before it begins. None is under way.