Human-cell-derived organoids recapitulate aspects of health and disease better than animal models. Their costs, logistics, and ethical profile allow deployment of extremely large-scale screens — opening experimental design spaces that were previously inaccessible.
Contrary to whole organisms, organoids have no endogenous homeostatic mechanisms to ensure robust developmental trajectories. This demands closed-loop control of oxygenation, temperature, nutrients, chemical and mechanical cues as they grow.
This control allows us to optimally bioengineer organoids toward our scientific questions — adding or reducing complexity at will. With the right information, the reproducibility problem dissolves into a parameter optimization problem.
Assume lab automation — from conveyor-belt-like fluidics handling intact specimens to humanoid robots — takes over specimen handling. AI scientists tailor workflows to optimally explore parameter spaces.
Automation → handling · AI → experiment design · ??? → information extraction
"How can we create scalable imaging pipelines that continuously image and real-time quantify information from 100–1,000s of organoids?"
Computational photography decoupled image quality from optical complexity. Can we apply the same paradigm to biomedical microscopy?
We are building AI-guided microscopy pipelines that combine scalable hardware with intelligent reconstruction — purpose-built for the organoid era.
Functional, structural & molecular imaging across resolutions
Exploiting light–tissue interaction for depth & contrast
Traditional & synthetic biology reporters matched to modality
Reconstruction with uncertainty quantification & validation