Ecological Modeling with Categories
Posted by John Baez
One of my dreams is finally coming true: ecological modeling with the help of category theory!
Topos Research UK, run by my friends and colleagues, will be providing computational support to ecological modelers using CatColab, their modeling software based on double categories. And they’re hiring a postdoc! I almost want to apply myself.
You can see details, and apply, here. The job starts now, essentially. The location is their wonderful office in Oxford, UK. It’s a full-time 2-year job with a salary of £47,665–52,430 per year. It’s in-person, but a transition period may be negotiated.
You should have, or soon have, a PhD in, or have made significant contributions to, mathematics or its adjacent fields (such as computer science, statistics, etc.), with a focus in at least one of the following areas:
- category theory,
- categorical statistics,
- (probabilistic) graphical models,
- systems dynamics,
- decision analysis,
- ecological modelling.
Before applying, I recommend that you read Tim Hosgood’s blog post about this job, since you’ll be working with him. I’ll quote a little:
To understand what sort of work this position entails, I think it’s helpful to look back at some of the work that people at Topos, and our collaborators, have been doing. Over the past year or so, I’ve been working as part of the CatColab team, looking at how we can take some of the foundational research work happening at Topos on declarative formal modelling (Lambert and Patterson 2024; Libkind and Myers 2025) and embed it into use-case contexts. Historically, a lot of this work has been contextualised within public health and epidemiology: AlgebraicPetri.jl (Libkind et al. 2022) is a core example, as is the AIM workshop that we ran at the start of this year with the US NSF Center for Analysis and Prediction of Pandemic Expansion (APPEX). More recently, we’ve been working with IDEMS and CASAS on software for modelling complex systems arising in the study of invasive species (Simmons et al. 2025).
There’s more, so read his whole post.