Founder and scientist
MD · PhD · Mathematician · Inventor · Deep-tech founder
Building technology for the programmable brain, and measuring intelligence in brains and in machines.
I am an MD PhD, mathematician, AI researcher and neurotechnology founder. I build systems at the intersection of neuroscience, AI, medical robotics and human enhancement, connecting measurement of the brain with intervention and verifying the individual response.
Science
Publications ↓Measurable intelligence, brain dynamics, learning systems and verifiable AI; brain imaging and medical robotics.
Business
Adapta ↓Brain–computer interfaces, programmable neuroplasticity, medical robotics and AI-native discovery.
Science
My research asks whether intelligence, consciousness and truth can be read from the organization of a system rather than from its outputs, and builds the instruments to do so: for the brain, for trained networks and for AI agents.
Measurable intelligence
A hierarchy of AI levels defined by the intrinsic dimension, topology and complexity of the data an intelligence generates, not by task performance; and the New Cognitive Paradigm, which moves from imitation to structural measurability.
Brain dynamics
BOID, an interpretable measure of the active oscillatory degrees of freedom in brain activity; with it, a multimodal prognostic model reached AUC 0.81 in 575 post-cardiac-arrest patients. Multichannel singular spectrum analysis estimates oscillatory sources and localizes attenuating oscillations.
Learning systems
The same instruments applied to learning machines: learning creates functional regime switching, a measurable signature shared by brains and trained networks.
Verifiable AI and agents
A legal verifiability compiler with a rule/judgment firewall, in which the model proposes and a deterministic compiler disposes; and when heterogeneous language-model agents do and do not reach Nash equilibrium in classic games.
Imaging and robotics
VeNet, a lightweight network for brain-vessel segmentation; an expert-annotated MRA dataset of the Circle of Willis; a robotic motion-copying system and a personalized vascular simulator for endovascular neurosurgery.
EEG and BCI methods
Hankel-embedding methods for counting active oscillatory components in EEG and for real-time filtering and classification on edge BCI systems; a topological framework for classifying time series.
Publications
A selection from more than forty papers, from cell biology and immunology to EEG, brain imaging and AI. Preprints with their DOIs are on AICumene Research.
* Co-author as a member of the European DLB Consortium; her name was omitted from the published author list in error.
Full list and current counts: Google Scholar · ORCID · OpenAlex.
Business
From dry-electrode hardware to AI-native discovery: measurement of the brain connected to intervention, and the individual response verified.
Current company
Adapta is being formed in the United States through the consolidation of technology and operations developed by NeuroSputnik and Nimbus BCI. It brings the whole loop of programmable neuroplasticity under one roof:
Brain–computer interfaces
Proprietary dry-gold-electrode systems with real-time acquisition, wireless connectivity and personalized neurofeedback, used by paying customers for monitoring, research and neuroplasticity programs.
Programmable neuroplasticity
Individualized interventions designed from multimodal data and checked against whether they actually change the individual brain; in an exploratory longitudinal study 12 of 12 participants showed measurable regional increases in cortical thickness.
Medical robotics
An endovascular robotic neurosurgery system and its AI-assisted navigation stack: vascular imaging pipelines, datasets and VeNet.
AI-native discovery
Computational screening and experimental validation of compounds and pro-cognitive formulations, linking biological activity with measurable cognitive and brain-response data.
Personalized BCI software
Adaptive EEG algorithms, real-time signal-processing pipelines and online personalization of individual brain-signal models.
AICumene
Non-invasive, intelligent brain–computer interfaces and an open preprint site for the science behind them.
Recognition
Why the brain
I started in mathematics and medicine because I wanted to understand complex biological systems rigorously. Measuring the brain turned out not to be enough: the real task is to connect measurement with intervention and verify the individual response. My long-term goal is to make brain adaptation measurable, personal and eventually programmable.
Contact
Open to researchers, clinicians, investors and founders in neurotechnology, AI, brain longevity, medical robotics and human–AI systems.