The next generation automatic sleep scoring system is here...

Unmatched reliability, super-fast turnaround time, future ready...

Test Drive Find Out More

One night of sleep record takes anywhere between 1.5 to 2.5 hours to score by an expert technologist. What if you could get hundreds of such records scored at expert level accuracy and higher reliability in seconds. In this world of ever increasing data - time and speed is everything!

Get Started!

At Your Service

Accurate and Reliable

Validated on over 11,000 hours of sleep records, ours systems are as accurate and more reliable than expert scorers (Patanaik et al., 2018, Sleep doi:10.1093/sleep/zsy041).

Real-time Scoring

Z3Score allows cutting-edge experiments that rely on real-time sleep stage information, like targetted memory reactivation, closed loop acoustic stimulation and lucid dreaming accessible to more researchers. Learn more about our pipeline.


We offer our service as a RESTful API, so you can integrate our scorers into any software or hardware platform. Read API documentation.

Become An Early Adopter

We are looking for early adopters of our technology, who are willing to work with us in the product development phase. You are welcome to test drive our prototype. Send your request to


Rapid Quality Checks.

Each sleep scored epoch also has an associated confidence. A score between 0 and 10, you can set your confidence threshold to get your desired level of quality. Figure below shows the percentage of data that is marked for review (bottom x-axis) for a corresponding confidence threshold value (top x-axis) and the automatic vs expert inter-rater reliability value for data that will not be reviewed (y-axis) for a representative dataset (more details will be published soon).

Open Source

We are a strong believer of open source. Go ahead and start scoring your PSG data using our open source MATLAB auto scoring client based on FASST (fMRI Artefact rejection and Sleep Scoring Toolbox). Or you can read our API documentation here. Or download our library to create compressed feature set files and streams from raw PSG data in MATLAB or Python.

Comparison of best available computer-assisted scoring software with Z3Score

Our system represents a significant improvement compared to the current state of the art and can bring considerable efficiency to your laboratories' workflow and output.

Developed by experts in the field

We are NeuroBit Technologies, a group of experts in neuroscience, machine learning, signal processing and bio-electronics with a vision to enhance human cognition through integration of smart immersive technology with better physiological data analytics.

Dr. Amiya Patanaik


is an expert in machine learning and statistical modeling with applications to computational neuroscience and human factors. He has worked in the area of close loop acoustic stimulations to boost sleep and holds a patent to do real-time sleep staging at expert human level accuracy. Download CV.



is an expert in hardware design and bio-electronics. With a background in experimental physics, he has worked on cutting edge products that use physiological data to augment human abilities.

Dr. Michael Chee

scientific advisor

is the Director of the Centre for Cognitive Neuroscience and Principal Investigator of the Cognitive Neuroscience Laboratory at Duke-NUS Medical School. The Z3Score framework was developed at his Laboratory after extensive R&D spanning across 3 years. With over 180 peer reviewed publications and over 8000 citations, he is a leading expert in the field of sleep and cognition. More Info...

See it in action.

Free Test Drive

If you are interested in test driving our prototype, please send in your request to

Get in touch

We use patent pending, state of the art Machine Learning algorithms instead of hard coded rules. Which means our system adapts and improves with more and more data. If you are interested in investing or licensing our technology, please get in touch. Send in your queries to


Powered By: