QURI is a nonprofit dedicated to dramatically improving estimation and forecasting using technology. We do this by developing software and doing research. We’re currently focused on scaling estimations regarding philanthropic resources for the benefit of the big-picture future of sentient life.
Right now QURI is made up of Ozzie Gooen and Nuño Sempere, plus multiple contractors. We’re looking to add two to three full-time equivalents in the next 12 months.
About the role
We’ve built an early version of Squiggle, a new programming language for estimation. Squiggle is somewhat of a mix between Guesstimate, Mathematica, and simple probabilistic programming languages.
It will take a lot of work to make Squiggle “production grade” and usable for many use cases. We’re looking for someone who could lead its development going forward.
Examples of what you will be doing
- Define and advance a language spec and a strong standard library.
- Flesh out programming language functionality and infrastructure.
- Build caching mechanisms for Monte Carlo simulations.
- Build functionality to handle frequently-updating imports (i.e. probabilistic models that reference other probabilistic models by different authors).
- Significant experience in programming language design, probabilistic programming, and/or scientific programming. It’s okay not to be great at all of these, but you should be robust in at least one or two.
- Be comfortable working with others on key decisions and functionality. For instance, we want someone who can make steady progress from the existing code base, as opposed to attempting first to rewrite it all from scratch.
- A passion for improving estimation and/or greater epistemic infrastructure.
Berkeley is preferred, but remote can work for particularly strong candidates.
$80,000 to $140,000, depending on experience.