Runtime
Biological systems become executable, testable, and versioned
A Computational Operating System for Biology
Living Systems OS
A foundational runtime to model, simulate, and steer living systems. LS-OS turns biology from static data into executable software.
Built by Kalman Labs.
Prediction
Trajectory-level forecasts across time, not one-time labels
Control
Simulate what-if interventions before expensive real-world decisions
Why LS-OS
Biology Has Data. It Needs an Operating System.
Today
- Models are fragmented and hard to execute end-to-end
- Predictions often fail when treatments change
- Biology is mostly analyzed, rarely run like software
With LS-OS
- Hybrid AI + mechanistic models inside one runtime
- Perturbation-aware simulation for real treatment scenarios
- Reproducible model lifecycle from data to deployment
Initial Wedge
Executable Tumor Dynamics in EGFR-mutant NSCLC
Start narrow. Solve deeply. We begin with EGFR-mutant NSCLC and EGFR TKIs to predict tumor response over time and anticipate resistance under different treatment strategies.
Cancer
EGFR-mutant non-small cell lung cancer
Therapy
Erlotinib, Gefitinib, Osimertinib
Data
Bulk RNA-seq + tumor volume trajectories
MVP
Aspirational, Practical, Measurable
Build
Create an executable tumor dynamics model with drug perturbation terms.
Train
Fit on patient cohorts using RNA-seq and longitudinal outcomes.
Validate
Beat Cox and black-box deep learning on held-out response trajectories.
Deploy
Support therapy ranking and resistance-aware trial decisions.
Collaborate in Research
Work With Kalman Labs on LS-OS
Cohort Studies
Joint retrospective analyses on NSCLC cohorts to test trajectory predictions.
Method Co-Development
Partner on model design, validation endpoints, and translational workflows.
Pilot Programs
Run practical pilots for treatment ranking and resistance-aware planning.