Build powerful quantitative models that trade and accelerate scientific discovery.

tl;dr: come build powerful quantitative models that trade and accelerate scientific discovery.

Historically, domain-specific models outperformed general models. New large models outperform across domains.

Zoa Research trains cross-domain forecasting engines. We deploy as (1) proprietary capital—algorithmic trading; (2) Forecast-as-a-Service—an API and console.

Motivation

Smart people should do important things. Too many of them trade instead. We would know—Sam worked at Jane Street for three years.

But there’s good news. Big models are now outperforming small models. Big models can trade, but also forecast supply chain volatility. Improve scientific priors. Forecasting generalizes.

And forecasting is a critical measure of scientific progress; science is, as Ian Hacking writes, the taming of chance.

About You

We’re looking for people with ML engineering skills or research experience. You’d be joining a highly motivated team of 3.

You may be a good fit if you:

• Have empirical AI/ML research experience

• Are good at math

• Are good at physics

• Quant trade

We are a fully in-person company, must be willing to relocate to NYC. Able to sponsor TN or E3 visas.

Hardest problems

• Making it fast

Stack

• Frontend: Next.js / React / Tailwind
• Backend: Next.js / Python
• Infrastructure: AWS