I’m a Scientist II in Computational Biology at Altos Labs, developing probabilistic machine learning methods and biophysical models to understand biological systems.
My research spans single-cell genomics, evolutionary dynamics, and gene regulation. I combine statistical physics, generative modeling, and Bayesian inference to ask: what can we reliably learn from noisy biological measurements?
I move between mechanistic models, statistical inference, and deep learning according to the biological question. My experience performing experiments—including fluorescence microscopy—shapes how I model measurements, test predictions, and interpret uncertainty.
Statistical physics and dynamical systems provide a foundation for connecting molecular mechanisms to cellular behavior. Probabilistic machine learning helps me bring that understanding to complex, high-dimensional data.
I build scientific software that makes these methods usable: explicit assumptions, reusable components, reproducible analyses, and documentation that connects the code to the science.
- BarBay.jl — Bayesian inference of relative fitness from high-throughput DNA barcode competition assays, with uncertainty quantification and models for multiple environments and experimental replicates. Documentation
- AutoEncoderToolkit.jl — Tools for training variational autoencoders and their extensions, including geometric analysis of learned latent spaces. Documentation
- SCRIBE — Probabilistic modeling of single-cell RNA-seq data with explicit treatment of measurement noise and uncertainty, using GPU-accelerated inference in JAX and NumPyro. Source code is currently private; publication in preparation.
Previously, I was a Schmidt Science Fellow and postdoctoral researcher at Stanford, working with Dmitri Petrov and collaborating with Madhav Mani at Northwestern University. My work focused on Bayesian fitness inference and geometry-aware representation learning for evolutionary landscapes.
I earned my PhD in Biochemistry and Molecular Biophysics at Caltech with Rob Phillips, combining statistical physics, information theory, and experiments to study how genetic circuits process environmental signals.