See the STRONGR
recommendation layer in action.

Real data. Explainable
recommendation intelligence.

This capstone model demonstrates the recommendation-intelligence layer of STRONGR OS using real public interaction data. Adjust the context, see how the score responds, and understand the factors influencing the result.

OPEN LIVE STRONGR OS CAPSTONE MODEL

PARITY EVIDENCE

One model.
Two execution environments.

100 real DEV rows tested8.326672684688674e-17

maximum sklearn/browser difference

Acceptance requirement≤ 1e-9PASS

WHAT THE LIVE MODEL SHOWS

Predicted Engagement Score

A browser-calculated model estimate, not a guarantee of an outcome.

Dataset Baseline

Context for the score against the official development positive prevalence.

Relative Engagement Index

A relative comparison, not a classification threshold or business rule.

Local Contributors & Governance

Active model arithmetic and boundaries—not causal explanations.

ACCEPTED INTERFACE EVIDENCE

Accepted public STRONGR OS model interface showing the Typical engagement context example: 4.01 percent predicted engagement score, 4.06 percent development baseline, 0.99 times relative engagement index, Near Baseline, and local model contributors. — Reviewer-approved public model — accepted default prediction example.