AI Research Scientist

Job Title: AI Research Scientist Location: Culver City, CA or Remote Employment Type: Full-Time About Kernel At Kernel, we are defining the future of brain health by making neuromeasurement accessible, actionable, and rigorously evidence-based. Our non-invasive Flow headsets pair TD-fNIRS with EEG to capture high-resolution functional brain data at scale. Today, we are building the largest neuroimaging dataset in history: 100,000+ hours of continuous, whole-head, multi-modal brain activity fueling breakthroughs in precision medicine, biomarker discovery, and AI. We are a passionate team of neuroscientists, engineers, and builders. If you want to solve challenges that expand our understanding of human cognition and health, we’d love to have you join us. About the Role We are hiring an AI Research Scientist to contribute to building a foundation model from our data, addressing limitations of traditional methods and reliance on heuristic assumptions to unlock new brain activity insights, revolutionizing brain function analysis and diagnostic accuracy. Responsibilities

• Design and implement specialized foundation model architectures for TD-fNIRS.
• Train and validate models using Kernel’s datasets, potentially supplemented with other datasets.
• Evaluate the usefulness of the model by leveraging it to help solve a subset of the problems that Kernel is working on: brain age prediction; cognitive scoring across the lifespan; prediction of treatment outcome for depression patients; classification of healthy controls vs mild cognitive impairment in an older population.
• Depending on the outcomes, there may be an opportunity to publish research findings in peer-reviewed journals and conferences.

Minimum Qualifications

• Strong background in machine learning, deep learning, and statistical analysis.
• Solid understanding of modern transformer architectures and training techniques.
• Proficiency in Python and relevant libraries (TensorFlow, PyTorch).
• Excellent written and verbal communication skills.
• Published research in academic journals or conferences in the fields of Computer Science, Neuroscience or Biomedical Engineering (PhD is not required).

Preferred Qualifications

• Experience building novel foundation model architectures, pretraining objectives, and data pipelines.
• Experience of foundation model experimental design - checkpointing, fine-tuning, establishing scaling behaviour.
• Experience of large scale data engineering and distributed/parallelised model training.
• Deep familiarity with the recent literature applying transformer models and deep learning to neuroimaging data.
• Experience with neuroimaging data analysis (fNIRS, EEG, MRI).
• Excellent data munging and visualization skills.
• Fluency with time series signal processing techniques.

Compensation Salary Range: $250–300k per year + equity + benefits