My name is Josh. I'm a generalist that's worked with a ton of different domains. If my specialty is anything it's translating whiteboard math into fast, usable software. Julia is my language of choice, and every project I build is driven by two priorities: performance and usability.
-
I'm the Co-Founder and Director of Software Engineering at Rallypoint One, a small business focused on open-source scientific computing, AI/ML, and digital twins for government R&D.
-
I do a lot of open source development.
-
I am the creator and main contributor to the Julia for Data Science newsletter.
-
I have a PhD in Statistics from NC State. My research into on-line algorithms for statistics led to the development of the OnlineStats Julia package.
I lead software engineering at RallypointOne, building Julia-based simulation and hybrid physics/ML models for federal R&D:
- ARID (NSF ASCEND Engine): Asset Resilience through Intelligent Digital Twins — systems integration and translation across a portfolio of researchers, startups, utilities, and government partners working on wildfire/water/power resilience in the Colorado–Wyoming region. Continues the wildfire digital-twin work started under HEATMAPS: GPU-accelerated simulation and continuous surrogate environment fields for fire decision support.
- MICROCAST (Navy SBIR): Hybrid forecasting framework for naval operations integrating SpeedyWeather, Oceananigans, Breeze, NumericalEarth, as well as numerical weather prediction tools from outside the Julia ecosystem.
- IRAM (NASA SBIR Ignite): Real-time hybrid physics/ML modeling of metal additive manufacturing — universal differential equations updated live from sensor data for defect prediction and closed-loop control.
- MIHSO (Metis Foundation): A multiscale cardiac digital twin for hemorrhagic shock and ischemia–reperfusion, built on the Julia SciML stack — coupling cell-scale electrophysiology, contraction, and mitochondrial energetics up to whole-heart simulation.
- ITCF (Navy STTR, with Colorado State): Integrated Tropical Cyclone Forecasting Suite — modernizing the Navy's legacy ATCF into a database-backed, API-driven architecture while preserving deck-file compatibility.
- OnlineStats.jl: Single-pass algorithms for statistics and data viz with big data.
- AverageShiftedHistograms.jl: Lightning-fast density estimation in Julia.
- PlotlyLight.jl: Lightweight, easy-to-use wrapper of the fantastic Plotly.js plotting library.
- XML.jl: An XML reader/writer in pure Julia that performs comparably to implementations in C.
- Cobweb.jl: An HTML reader/writer in pure Julia that's fast and easy to use.
- Wildfires.jl: Wildfire modeling and simulation in Julia.
- Elmfire.jl: Julia implementation of the ELMFIRE fire propagation model.
- GeoSurrogates.jl: Surrogate function representations of geospatial data.
- WildfireData.jl: Unified Julia interface for federal wildfire data (LANDFIRE, NIFC, GOES-16, HRRR).
- DepotDelivery.jl: Create a standalone Julia depot for air-gapped deployments.
- Terse.jl: Shorthand for defining full type hierarchies.
- TokenIterators.jl: Fast framework for writing lexers/tokenizers.
- XKCD.jl: Retrieve image and metadata from the XKCD webcomic.
- One Click Tuner: iOS app with a chromatic musical instrument tuner.
And much more... browse my repos or RallypointOne's.