Hemispheric
Category: AI in Healthcare
Hemispheric is a NeuroAI company building a frontier foundation model (Descartes) to decode human brain activity from non-invasive EEG for diagnosing and treating neurological and psychiatric disorders. Hemispheric was founded in 2019. The company is led by Hagai Lalazar. Based in Tel Aviv, Israel. Team size: 101-500. Total funding raised: $52M. Latest round: Series A. Key investors include Hanaco Ventures, Arkin Capital, OurCrowd, Artofin VC, OneMind/Awareness Capital, Protocol Labs, L Catterton, Howard Morgan, Naomi Azrieli, Yasmin Lukatz, Scott Belsky.
- Founded
- 2019
- Headquarters
- Tel Aviv, Israel
- Team size
- 101-500
- Total funding
- $52M
Value proposition
First frontier NeuroAI foundation model that decodes non-invasive brain activity to provide objective, quantitative brain health diagnostics — akin to a blood test for the brain — replacing subjective questionnaires and behavioral observations
Products and solutions
Descartes — a 6-billion-parameter NeuroAI foundation model trained on 250,000+ hours of multimodal EEG and behavioral data from 100,000+ participants, dry-electrode EEG headset for 15-minute non-invasive brain assessment, AI-powered diagnostic platform for PTSD, mild traumatic brain injury (mTBI), depression, anxiety, schizophrenia, and Alzheimer's disease
Unique value
The only NeuroAI company combining a 6B-parameter foundation model trained on 250K+ hours of proprietary brain data with a non-invasive, low-cost EEG headset to deliver objective brain diagnostics at scale, built by the co-inventor of Apple's FaceID and a computational neuroscientist
Target customer
Healthcare organizations, hospitals, mental health clinics, psychologists' offices, pharmaceutical companies, government agencies, and researchers
Industries served
Healthcare (neurology, psychiatry, mental health), Pharmaceuticals (clinical trials, drug development), Government (defense, veterans' health), Research
Technology advantage
6-billion-parameter Descartes NeuroAI foundation model trained on one of the world's largest proprietary brain-activity datasets (250K+ hours, 100K+ participants); proprietary data collection labs and AI training infrastructure built specifically for neural data; dry-electrode EEG headset requiring no surgery or invasive procedures; team of 112 with expertise spanning computational neuroscience, AI, and medical imaging; co-founders with Apple FaceID/Vision Pro and computational neuroscience pedigree
How they differentiate
Unlike invasive brain-computer interfaces (Neuralink, Synchron) that require surgery, Hemispheric uses a non-invasive dry-electrode EEG headset. Unlike traditional EEG diagnostic tools, Hemispheric has built a frontier foundation model (Descartes) trained on massive proprietary data, enabling it to generalize across individuals and conditions. The company's bet is that AI models trained on large-scale data, not better hardware, will unlock precision brain diagnostics.
Main competitors
Synchron (invasive BCI with Chiral foundation model), Neuralink (invasive BCI), Kernel (non-invasive EEG/neuroimaging), BrainKey (AI brain scan analysis), Coherence Neuro (cancer-detecting brain implants)
Key partnerships
Working with U.S. FDA Center for Devices and Radiological Health, partnerships with government agencies and pharmaceutical companies for clinical applications
Notable customers
Government agencies (undisclosed), pharmaceutical companies (undisclosed), healthcare organizations (undisclosed)
Major milestones
2019-2020: Company founded by Hagai Lalazar and Gidi Littwin, 2020-2026: Six years in stealth building Descartes model, collecting data from 100K+ participants across Asia, Tel Aviv, and Boston, Jul 2026: Emerged from stealth with $52M funding and launch of Descartes, 2027 (planned): Submit first PTSD diagnostic product to FDA for approval, 2027 (planned): Public rollout of PTSD product
Growth metrics
112 employees; 250,000+ hours of brain data; 100,000+ research participants; 6-billion-parameter model
Market positioning
First-mover in the frontier NeuroAI foundation model space for non-invasive brain diagnostics; positioned as the "foundation model for the human brain" analogous to LLMs for text or vision models for images
Geographic focus
Global (HQ in Israel, operations in US and Asia, pursuing FDA approval in US and regulatory approvals in Europe)
Patents and IP
Proprietary data collection infrastructure, model training systems designed for neural data, and dry-electrode EEG headset technology (specific patents not publicly detailed)
About Hagai Lalazar
Computational neuroscientist with 25+ years in scientific research; Post-doctoral Researcher at Center for Theoretical Neuroscience, Columbia University; PhD from Hebrew University of Jerusalem
Official website: https://www.hemispheric.ai