Hi, I'm Ben
I work on genomic language models at Radical Numerics. In my spare time, I write about quantitative finance, mountain biking, agentic coding, and lots more.
Previously, I worked on logic-gated cancer therapeutics at Pattern Biosciences. Before that, computational infrastructure and data tooling at Deep Origin, and I led the bioinformatics team at Loyal, a clinical-stage biotech dedicated to extending the lifespan and healthspan of dogs.
What seems like ages ago, I did my PhD in Dr. Ami Bhatt's lab at Stanford, where I developed and used advanced genomic techniques to study the human microbiome. I'm proud to say I characterized the first examples of bacteria transmitted between the gut microbiome of adults, and of phages transmitted from the gut microbiome of a mother to that of her infant.
Latest writing
All posts →-
Deep learning statistical arbitrage part 1 - replication
Reproducing the headline results from "Deep learning statistical arbitrage" (Guijarro-Ordonez et al. 2022) on my own data, with a python implementation you can use.
-
Revisiting VIX Call hedging - 2025 data update and new strategies
Five years on: revisiting OTM VIX-call tail hedges with a 2025 data extension - equity curves, risk/return frontier, and timing luck reduction, reproduced in Python.
-
Launching Radical Numerics
Radical Numerics emerges from stealth with $50M of seed funding and state of the art genomic language models.
-
Reducing RNA-seq batch effects by re-aligning TCGA and GTEx
Batch effects are a constant pain when comparing RNA-seq data from different sources, but re-processing with consistent bioinformatics tools can reduce much of the effect.
-
s3stasher simplifies using AWS S3 in python
s3stasher is a python package that transparently handles files in s3. It also maintains a local cache of recent files for offline use.
-
Fixing a Milli-q purifier for 99% off
Milli-q water purifiers are too smart for their own good, but splicing in a large resistor can bypass some of the restrictions.