Technology Focus

We create methods that can recover cardiovascular state from indirect or peripheral measurements using mechanistic models, waveform analysis, and uncertainty-aware inference. We leverage hemodynamic modeling, digital twin approaches, Bayesian methods, and neural surrogates of physical systems.

We develop computational methods for optical, wearable, and microfluidic sensor design under realistic variability. This includes simulation-driven design, robust optimization, and methods that account for conditions that often break the average-case sensor performance. We have expertise in light-tissue interaction modeling, multi-fidelity simulation workflows, and designing methods to support real-world validation and transition. 

We build tools that support technical de-risking before expensive experimental or clinical programs are required. This includes synthetic data pipelines, surrogate models, evaluation assets, and validation-oriented workflows. We have expertise in synthetic data generation for biosensing, virtual patient/population models, and developing surrogate models of physical systems for fast iterations.

We also have expertise in computer vision systems for medical and surgical use cases, particularly when robust inference depends on combining video with other sensor data and physiological or clinical data. This includes perception systems that support monitoring, quality assurance, workflow understanding, and decision-support tools.