Unlocking human potential through eye tracking

Our cutting edge algorithms are built on at least 15 years of in-house cognitive science and vision research demonstrating that eye movements provide a measurable window into human thought processes. We are a non-invasive approach, placing high value on data privacy.

A key part of our work involves pupillometry, which is highly sensitive to changes in brightness and vergence. Our core expertise lies in separating genuine cognitive signals-such as mental effort and attentional allocation-from environmental noise like lighting conditions or display luminance. By analyzing fixation duration, saccadic patterns, pupil dynamics, and visual search efficiency, we extract quantitative algorithms that directly reflect underlying cognitive functions. These algorithms are not subjective interpretations; they are derived from well-established links between oculomotor behavior and neural activity. Because gaze behavior is tightly coupled with cognition, it enables an objective, temporally precise, and real-time assessment of attention, working memory, processing speed, and executive control.

Importantly, our approach prioritizes flexibility in sampling rates and is not constrained to high-frequency data. This makes our models particularly adaptable and well-suited for integration into power-sensitive devices such as smartglasses, where energy efficiency is critical. At the base of it, using advanced statistical modeling and machine learning, we are able to transform raw eye tracking data into structured, interpretable cognitive intelligence. Our system identifies reproducible patterns associated with mental effort, attentional allocation, and perceptual engagement, but rather than relying on black-box AI, we follow a hybrid strategy that combines rule-based models with machine learning. This allows us to maintain interpretability while leveraging machine learning to enhance accuracy and robustness.

Scientific rigor defines our development process.

Our research team conducts systematic in-house experimentation using controlled task paradigms designed to isolate specific cognitive functions. Each algorithm is tested for sensitivity, robustness, and reproducibility across diverse conditions and participant groups. This ensures that our metrics respond accurately and predictably to changes in cognitive demand and perform reliably in real-world applications.

Overtime, we continuously refine our models using experimental datasets, benchmark comparisons, and complementary neurophysiological validation frameworks, including EEG-based paradigms. This cross-validation confirms alignment between physiological signals and established cognitive constructs.

Through this iterative process, we ensure our algorithms remain scientifically sound, technologically advanced, and grounded in empirical evidence.