Running AI like neurons, on neurons.
At Neuraffica, the mechanistic understanding of neuronal computation drives everything we build. Through targeted experiments, we reveal how biological circuits process signals and generate intelligence: not just their anatomical structure, but their functional logic. We translate these principles into neuro-inspired algorithms that mirror cortical computation and biological hardware platforms where living neurons accelerate AI training.
Intelligence grounded in biological reality, not infrastructure constraints.
We envision a future where AI systems are efficient by design, interpretable at their core, and naturally aligned—built on a mechanistic understanding of how biological computation actually works rather than optimized for existing hardware.
Neuraffica decodes biological intelligence to engineer the next generation of AI.
We place living neurons at the center of our approach. Using breakthrough neurotechnology, we map not just brain structure but the functional dynamics and operational rules that generate computation.
As an AI-native company, we augment human work with dedicated agentic infrastructure to eliminate cross-disciplinary collaboration bottlenecks.
Two paths to energy-efficient AI. Each reinforces the other.
Running AI On Neurons
Living neurons are our minimal intelligent system—learning with a fraction of the energy silicon requires. Others treat bio-hybrid platforms as black boxes. We make them designable: mapping how network structure creates computational properties, then engineering those properties on purpose. Our goal: energy-efficient AI training.
Running AI Like Neurons
We build alternative AI architectures inspired by mechanistic neuroscience and physics. Models that route signals the way neurons do: sparse, structured, and efficient by design—not optimized for hardware, but for intelligence.
Every experiment produces reusable computational rules. This growing library of mechanisms powers our architectures and extends to neuromorphic computing, chip design, and intelligent systems.
Our name tells our story. Neur anchors us in neuroscience—the mechanistic study of how biological circuits compute. Ffica captures the dynamic efficiency we're engineering: efficacy in computation, traffic of signals flowing through networks. Neurons are evolution's solution to energy-efficient intelligence. We're decoding how.
Neuraffica is built on a simple thesis: the next AI breakthrough requires neuroscientists who deeply understand AI, engineers who think like biologists, and seamless collaboration between fields that rarely intersect.
We are a stealth startup with a co-founding team that spans the full stack—from running biological lab and building neurotechnology to training LLMs from scratch and deploying large-scale agentic systems. Our expertise covers experimental neuroscience, computational neuroscience, machine learning, electrical engineering, optics, and molecular biology—each of us bringing irreplaceable experience from years of specialized research and production environments. Ex-UC Berkeley. Ex-Google. Ex-Spotify. Perfect blend of academic and industry settings. Ready to build day one.
Insights on AI, neuroscience, and our vision for the future.
Coming soon. Stay tuned for updates on our research and progress.
We're looking for exceptional talent across neuroscience, engineering, biology, and AI! If you share our belief that understanding biological intelligence is the key to breakthrough AI, we want to hear from you.
Bay Area, California
Don't see a fit? Reach out anyway.