Hardware
Components
Hardware
Components
Hardware
Components
+
+
01
01
Wearable
Recording Device
Wearable
Recording Device
Pendant Pin
Pendant Pin
02
02
Single EEG
Electrode
Single EEG
Electrode
Forehead
Forehead
03
03
Magnetic
Pin
Magnetic
Pin
Magnetic
Pin
Clothes Attachment
Clothes Attachment
Clothes Attachment
Interaction
architecture
Interaction
architecture
Interaction
architecture
+
Click each interaction to learn more and watch scrappy demos:
Click each interaction to learn more and watch scrappy demos:
mind
click
+
the
"archivist"
+
face
blur
+
voice
blur
+
: a neural-first data capture system that preserves meaningful moments without interrupting presence. By combining continuous sensing with real-time neural activity monitoring, it selectively commits data only when a user’s internal intention is detected. This removes the need for manual capture, reducing cognitive load while ensuring contextual relevance. The result is a seamless bridge between lived experience and recorded memory.
mind
click
+
the
"archivist"
+
face
blur
+
voice
blur
+
: a neural-first data capture system that preserves meaningful moments without interrupting presence. By combining continuous sensing with real-time neural activity monitoring, it selectively commits data only when a user’s internal intention is detected. This removes the need for manual capture, reducing cognitive load while ensuring contextual relevance. The result is a seamless bridge between lived experience and recorded memory.
mind
click
+
the
"archivist"
+
face
blur
+
voice
blur
+
: a neural-first data capture system that preserves meaningful moments without interrupting presence. By combining continuous sensing with real-time neural activity monitoring, it selectively commits data only when a user’s internal intention is detected. This removes the need for manual capture, reducing cognitive load while ensuring contextual relevance. The result is a seamless bridge between lived experience and recorded memory.
Mind
Click
+
A thought-activated system that retroactively captures meaningful moments with semantic contextualisation.
A thought-activated system that retroactively captures meaningful moments with semantic contextualisation.
45s
45s
capturing
capturing
trigger
trigger
Real-time audio-visual buffering on-device, with app storage activated only through mental triggers.
Real-time audio-visual buffering on-device, with app storage activated only through mental triggers.
time
![](data:image/svg+xml,<svg display="block" role="presentation" viewBox="0 0 560 299" xmlns="http://www.w3.org/2000/svg"><path d="M 0 299 L 560 299 L 560 0 L 0 0" fill="transparent" height="299px" id="vL8SNY6l0" stroke-dasharray="8" stroke-linecap="butt" stroke-linejoin="miter" stroke-miterlimit="4" stroke-width="2" stroke="rgb(0, 85, 255)" width="560px"/></svg>)
+
45s video clip
45s video clip
45s audio clip
45s audio clip
contextual summary
contextual summary
LLM API
LLM API
![](data:image/svg+xml,<svg display="block" role="presentation" viewBox="0 0 100 101" xmlns="http://www.w3.org/2000/svg"><path d="M 0 101 L 361 101 L 361 0" fill="transparent" height="101px" id="VOUFReo75" stroke-dasharray="8" stroke-linecap="butt" stroke-linejoin="miter" stroke-miterlimit="4" stroke-width="1" stroke="rgb(0, 85, 255)" transform="translate(-261 0)" width="361px"/></svg>)
Continuously buffers real-time audio and video locally on-device in a retroactive rolling window. When an intentional mental trigger is detected, previous moments are instantly preserved, saving video, audio, and an LLM-generated contextual summary. This allows users to capture meaningful experiences after they happen, transforming thoughts into memories.
Continuously buffers real-time audio and video locally on-device in a retroactive rolling window. When an intentional mental trigger is detected, previous moments are instantly preserved, saving video, audio, and an LLM-generated contextual summary. This allows users to capture meaningful experiences after they happen, transforming thoughts into memories.
45s
capturing
trigger
Real-time audio-visual buffering on-device, with app storage activated only through mental triggers.
time
+
45s video clip
45s audio clip
contextual summary
+
![](data:image/svg+xml,<svg display="block" role="presentation" viewBox="0 0 95 70" xmlns="http://www.w3.org/2000/svg"><path d="M 0 70 L 95 70 L 95 0" fill="transparent" height="70px" id="L0I8gRFRM" stroke-dasharray="8" stroke-linecap="butt" stroke-linejoin="miter" stroke-miterlimit="4" stroke-width="1" stroke="rgb(0, 85, 255)" width="95px"/></svg>)
![](data:image/svg+xml,<svg display="block" role="presentation" viewBox="0 0 95 70" xmlns="http://www.w3.org/2000/svg"><path d="M 95 70 L 0 70 L 0 0" fill="transparent" height="70px" id="meIC74hiy" stroke-dasharray="8" stroke-linecap="butt" stroke-linejoin="miter" stroke-miterlimit="4" stroke-width="1" stroke="rgb(0, 85, 255)" width="95px"/></svg>)
LLM API
Mind
Click
Mind
Click
+
A thought-activated system that retroactively captures meaningful moments with semantic contextualisation.
45s video clip
45s audio clip
contextual summary
Continuously buffers real-time audio and video locally on-device in a retroactive rolling window. When an intentional mental trigger is detected, previous moments are instantly preserved, saving video, audio, and an LLM-generated contextual summary. This allows users to capture meaningful experiences after they happen, transforming thoughts into memories.
Continuously buffers real-time audio and video locally on-device in a retroactive rolling window. When an intentional mental trigger is detected, previous moments are instantly preserved, saving video, audio, and an LLM-generated contextual summary. This allows users to capture meaningful experiences after they happen, transforming thoughts into memories.
![](data:image/svg+xml,<svg display="block" role="presentation" viewBox="0 0 1 341" xmlns="http://www.w3.org/2000/svg"><path d="M 0 340.5 L 0 0" fill="transparent" height="340.5px" id="RetMcCAz2" stroke-dasharray="8" stroke-linecap="butt" stroke-linejoin="miter" stroke-miterlimit="4" stroke-width="2" stroke="rgb(0, 85, 255)" width="1px"/></svg>)
+
LLM API
Click on the crosses to explore hardware interactions and details.
Click on the crosses to explore hardware interactions and details.
Click on the crosses to explore hardware interactions and details.
Focus level
Focus level
Focus level
alpha waves (relaxation)
alpha waves (relaxation)
beta waves (alertness)
beta waves (alertness)
binary output
not focused
binary
output
focused
Focus level is determined by analysing real-time EEG signals, particularly the balance between alpha waves (relaxation) and beta waves (alertness). These patterns are translated into a binary output that identifies whether the user is focused or not, forming the system’s core cognitive detection layer.
Focus level is determined by analysing real-time EEG signals, particularly the balance between alpha waves (relaxation) and beta waves (alertness). These patterns are translated into a binary output that identifies whether the user is focused or not, forming the system’s core cognitive detection layer.
Focus level is determined by analysing real-time EEG signals, particularly the balance between alpha waves (relaxation) and beta waves (alertness). These patterns are translated into a binary output that identifies whether the user is focused or not, forming the system’s core cognitive detection layer.
Pre training
Pre training
Pre training
focus
level
focus
level
ML
classes
ML
classes
raw
EEG data
raw
EEG data
signal
processing
signal
processing
class
data
class
data
-
+
![](data:image/svg+xml,<svg display="block" role="presentation" viewBox="0 0 1908 1" xmlns="http://www.w3.org/2000/svg"><path d="M 0 0 L 1908 0" fill="transparent" height="1px" id="QP1aD6_q2" stroke-dasharray="8" stroke-linecap="butt" stroke-linejoin="miter" stroke-miterlimit="4" stroke-width="4" stroke="rgb(80, 80, 80)" width="1908px"/></svg>)
"idle|random"
"idle|random"
"capture the past"
"capture the past"
![](data:image/svg+xml,<svg display="block" role="presentation" viewBox="0 0 934 90" xmlns="http://www.w3.org/2000/svg"><path d="M 0 90 L 732 90 L 732 0 L 933.5 0" fill="transparent" height="90px" id="r2n04pVlF" stroke-dasharray="8" stroke-linecap="butt" stroke-linejoin="miter" stroke-miterlimit="4" stroke-width="2" stroke="rgb(80, 80, 80)" width="933.5px"/></svg>)
![](data:image/svg+xml,<svg display="block" role="presentation" viewBox="0 0 1908 1" xmlns="http://www.w3.org/2000/svg"><path d="M 0 0 L 1908 0" fill="transparent" height="1px" id="sNLFl49KI" stroke-dasharray="8" stroke-linecap="butt" stroke-linejoin="miter" stroke-miterlimit="4" stroke-width="3" stroke="rgb(80, 80, 80)" width="1908px"/></svg>)
2025-03 31T18:06:01.036586,-26.359169006347656,-3.3687331676483154,idle
2025-03 31T18:06:01.036586,-26.359169006347656,-3.3687331676483154,idle
2025-03-31T18:06:01.137660,-66.3968505859375,31.03977394104004,Capture
2025-03-31T18:06:01.137660,-66.3968505859375,31.03977394104004,Capture
Capture
Capture
Idle
Idle
The pretraining process introduces a novel method of teaching the system to recognise intentional thought patterns by pairing user-defined mental commands with EEG-derived focus states. Rather than relying on generic brain-computer interaction models, it creates a personalised cognitive framework where the model learns the specific neural signature of deliberate intention. This transforms raw EEG data into actionable thought-based commands, forming the foundation for reliable retroactive memory capture and significantly advancing practical, consumer-ready brain-computer interaction.
The pretraining process introduces a novel method of teaching the system to recognise intentional thought patterns by pairing user-defined mental commands with EEG-derived focus states. Rather than relying on generic brain-computer interaction models, it creates a personalised cognitive framework where the model learns the specific neural signature of deliberate intention. This transforms raw EEG data into actionable thought-based commands, forming the foundation for reliable retroactive memory capture and significantly advancing practical, consumer-ready brain-computer interaction.
FOCUS LEVEL > ML CLASSES > RAW EEG DATA > SIGNAL PROCESSING > CLASS DATA
The pretraining process introduces a novel method of teaching the system to recognise intentional thought patterns by pairing user-defined mental commands with EEG-derived focus states. Rather than relying on generic brain-computer interaction models, it creates a personalised cognitive framework where the model learns the specific neural signature of deliberate intention. This transforms raw EEG data into actionable thought-based commands, forming the foundation for reliable retroactive memory capture and significantly advancing practical, consumer-ready brain-computer interaction.
-
+
"idle|random"
"capture the past"
![](data:image/svg+xml,<svg display="block" role="presentation" viewBox="0 0 1 131" xmlns="http://www.w3.org/2000/svg"><path d="M 0 0 L 0 131" fill="rgb(119, 119, 119)" height="131px" id="jWuNHb0F0" stroke-dasharray="8" stroke-linecap="butt" stroke-linejoin="miter" stroke-miterlimit="4" stroke-width="3" stroke="rgb(119, 119, 119)" width="1px"/></svg>)
2025-03 31T18:06:01.036586,-26.359169006347656,-3.3687331676483154,idle
2025-03-31T18:06:01.137660,-66.3968505859375,31.03977394104004,Capture
Capture
Idle
![](data:image/svg+xml,<svg display="block" role="presentation" viewBox="0 0 95 254" xmlns="http://www.w3.org/2000/svg"><path d="M 95 254 L 95 160.5 L 0 160.5 L 0 0" fill="transparent" height="254px" id="PRSRFENS4" stroke-dasharray="8" stroke-linecap="butt" stroke-linejoin="miter" stroke-miterlimit="4" stroke-width="2" stroke="rgb(102, 102, 102)" width="95px"/></svg>)
Competitions & Collaborations
Competitions & Collaborations
Competitions & Collaborations