01 · What is established

The signal that has earned confidence.

  • People vary in postprandial responses, preferences, behaviors, and metabolic risk.
  • Wearables and at-home testing can create dense longitudinal data outside traditional clinical settings.
  • The current U.S. framework was not designed for services that combine foods, devices, algorithms, and health recommendations.

02 · What remains unsettled

The uncertainty is part of the answer.

  • Which response phenotypes reproduce across time and meaningfully improve decisions.
  • How algorithms should be validated when both the model and the user behavior change.
  • Where wellness guidance ends and regulated medical or diagnostic functionality begins.

03 · What this changes

From scientific landscape to decision architecture.

Berkwood evaluates the full evidence chain behind personalized systems, identifies the validation burden for the intended claim, and helps teams create frameworks that protect users while enabling responsible innovation.

Selected work

Read into the evidence.