Sapien Weekly Digest
What we’ve been up to:
We ran the first end to end verification workflow through Proof of Quality, a dataset of skin lesions. Here’s what changed and what comes next.
The App in Action during our test run
What we learned:
The goal was straightforward:
Run a real workflow end to end.
Measure how group consensus tracks against a reference set.
Turn that into a usable baseline for Proof of Quality.
The run gave us a working baseline for how the current consensus flow performs with general validators without subject matter expertise. We analyzed 154 scored items in total. Across the full set, group consensus matched the reference a majority of the time, with an interval of only ±7.1%, despite the lack of medical qualification.
What happens next:
Next up is external validation with expert reviewers from the medical field.
Same workflow.
Subject Matter Expert validators.
Production settings.
This next run will show how Proof of Quality performs when domain expertise enters the loop.
Build on Proof of Quality.
Ship AI systems with quality signals you can verify. Join the Proof of Quality waitlist and start building on Sapien and join the early adopters list for priority access when it goes live!
