Now in development
Imagine being understood.
Hypersigil is developing a brain–computer interface that is worn, not implanted, and calibrated to the way each person imagines.
The idea
People imagine differently.
Ask 10 people to imagine the ocean. Some see waves. Some hear surf. Some feel cold water. A few report no picture at all.
Psychologists have measured these differences for decades. Brain–computer interfaces have made little use of them. Our hypothesis is that an interface that knows how a person imagines can be calibrated more accurately to that person.
How it works
Measure the person first.
- 01
Measure.
Before any recording, a structured assessment measures how vividly a person imagines in eight senses: sight, sound, smell, taste, touch, movement, bodily sensation, and emotion.
- 02
Calibrate.
The results shape that person’s calibration tasks. While the person imagines a chosen concept in several senses, a headset records electroencephalography (EEG), the brain’s electrical activity measured at the scalp.
- 03
Test.
We test whether the recordings identify the concept the person imagined, and whether calibration designed around the person performs better than calibration that is not.
We have not yet demonstrated recognition of imagined concepts. That is the subject of our next experiment.
Where the work stands
Progress to date.
Imagery measure
Developed in Dylan Furdyk’s doctoral research at King’s College London and tested in a study of 165 participants. A paper is in preparation.
Calibration software
A Mac application presents the imagery tasks, checks signal quality, and records each session.
First recordings
A first session with an 8-channel headset was completed in November 2026. It tested the workflow and is not yet research data.
Research instrument
A 64-channel wireless instrument has been designed. Review and fabrication come next.
Open data
We have analyzed 14 public brain-recording datasets, collected by other research groups, to guide the design of our experiments.
The headset
Worn, not implanted.
The interface records from outside the head. There is no surgery and no implant, and the headset can be removed at any time.
The first instrument is built for research laboratories: a 64-electrode cap connected to a compact wireless unit. A slim headset is the form we are designing toward.
- No surgery
- No implants
- Removable
The science
A century-old signal, calibrated to one person.
Neurons in the outer layer of the brain produce small electrical signals. Electrodes on the scalp can record them. The method, EEG, has been in use since the 1920s.
EEG is painless and requires no surgery. The signal is faint, and it differs from one person to the next. Our research asks whether measuring how a person imagines makes that signal easier to interpret.
At a glance
- Form
- Worn on the head; removable
- Signal
- EEG, recorded at the scalp
- Calibration
- An imagery assessment in eight senses, then short recording tasks
- First product
- Calibration software for research laboratories
- Hardware
- 64-channel wireless research instrument, in development
- Status
- In development; concept recognition not yet demonstrated
- Patent
- Patent pending. U.S. provisional patent application filed 30 October 2026
Where this could lead
Long-term aims.
None of these has been demonstrated.
-
Search
Retrieve information by bringing a trained concept to mind.
-
Create
Direct drawing and design software with imagined concepts.
-
Communicate
Operate speech and communication aids for people who cannot easily speak or type.
Privacy and control
Designed around consent.
Brain recordings are personal data. We are designing the system to recognize only concepts a person has chosen to train, and to return no result when it is uncertain. We will ask for specific consent before any recording is used to improve a shared model. The headset can be taken off at any time.
For researchers
For research laboratories.
Our first product is calibration software for laboratories that study mental imagery or EEG decoding. It runs with compatible EEG headsets. If this is your field, write to info@hypersigil.org.
The team
Who is building it.
Dylan Furdyk
Chief executive
PhD candidate in psychology at King’s College London. Developed the imagery measure and leads experimental design.
Adam Majid
Chief technology officer
About 15 years in software engineering and product design. Designed the 64-channel instrument and leads hardware, software, and machine learning.
Questions, answered.
What is Hypersigil?
Hypersigil, Inc. is developing a noninvasive brain–computer interface: a headset that records electroencephalography (EEG), and software calibrated to the way each person imagines.
Is it a brain implant?
No. The headset is worn on the head and records from the scalp. There is no surgery and no implant.
How is it different from implanted interfaces?
Implanted interfaces place electrodes in or on the brain through surgery, and they record stronger signals. A worn interface records weaker signals, carries no surgical risk, and can be removed. Our research concerns how much a worn interface can gain from calibration to the individual.
How does it work?
We first measure how a person imagines in eight senses. We use that profile to design calibration tasks, record EEG while the person imagines chosen concepts, and test whether the recordings identify those concepts.
Can it read my thoughts?
No. The aim is to recognize specific concepts that a person has chosen to train and deliberately brings to mind. It is not designed to recover private or unprompted thoughts.
Does it work now?
Not yet. The imagery measure, the calibration software, and a first recording session exist. Recognition of imagined concepts is the subject of our next experiment.
Who is it for?
Research laboratories first. Later, people who need new ways to communicate, and eventually general use.
Is it a medical device?
No. It is a research system. It is not intended to diagnose, treat, or prevent any condition.
When can I try it?
No date is set. Join the waitlist by writing to info@hypersigil.org, and we will write when there is something to try.
Follow the work.
Join the waitlist and we will write when there is something to try.
Join the waitlistOr write to info@hypersigil.org. Read directly by the founders.