Our AI-powered Analog Cell Emulator, built on a Spiking Neural System, replicates the real-world complexity of human cells in an ultra-noisy (SNR<1), asynchronous, distributed-clock, non-linear environment — identifying risk early and shortening the path from candidate molecule to clinical impact.
Cells operate with inherent stochasticity, and function seamlessly across multiple, decentralized timescales dictated by local memory networks — such as epigenetic states and chromatin topology.
Digital neural networks require immense energy, processing power, and forced abstraction to approximate these non-linear, continuous-time dynamics. By designing our hardware in analog, our silicon physically mirrors the state changes and local memory networks of the cell — allowing for real-time, low-energy, highly accurate simulation.
Cellular noise isn’t a flaw to filter out — it’s part of the signal. Our platform emulates biology at its native, ultra-noisy operating point.
Local memory networks — epigenetic states, chromatin topology — run on decentralized clocks. Our spiking system mirrors that asynchrony directly.
Analog hardware physically mirrors cell-state changes, delivering real-time, low-energy simulation without the abstraction overhead of digital approximation.
A live look at how a candidate molecule moves through the emulation platform before it ever reaches a wet lab.
We’re a small, interdisciplinary team working across analog circuit design, computational biology, and machine learning.
View open roles