The only platform for real‑time, quality stress data.

Build emotionally intelligent AI using the world’s first accurate real-time stress measure.

Collaborating with advisors and researchers from

Problem

Stress data today is either unviable or unreliable

Labs are limited and unviable

Lab tests like cortisol are good for research but inaccurate for continuous measurement. They are also expensive, slow and incompatible with commercial use.

Wearable readings are wrong about half the time

In our Cambridge–Dresden study, the next best wearable score agreed with biomarker-confirmed stress 54.7% of the time.

Stress data today is commercially unsuitable

Building products around stress is limited by how little data there is, and how inaccurate it is.

Solution

CSR models unlock quality stress data in real‑time

Proprietary models, two granted patents

Our models label stress data using heuristic and physiological models we own the patents to.

Works on nearly any wearable

It uses sensors that consumer wearables already have, so there is no new hardware to wear.

Made for commercial use

Stress is measured in real time with zero ongoing cost, at 30x the resolution of the next best measure.

An accurate picture of chronic stress

The scale has 10 subcategories, which identify the best intervention for each individual and across whole organisations.

Cambridge–Dresden study, 2026

95.7% agreement with biomarker-confirmed stress

With the University of Cambridge and TU Dresden, we tested our wearable-derived measure on 113 people under free-living conditions. The next best wearable score was just 54.7%. Peer-reviewed paper submitted October 2026.

95.7%

Cambridge Stress Research measure

54.7%

Next best wearable score

Stress data is the key to artificial emotional intelligence

Stress data reveals our emotional state
General artificial intelligence needs our emotional state
Google and others spent millions cracking stress data
They failed because AI cannot label stress data on its own
Our patents label stress data in real time

Let’s collaborate

We invite academics to enquire about research projects that could benefit from our measure, and organisations that want to work with it.

Prefer email? info@cambridgestressresearch.com