**About Terranox**
We're the first AI-powered uranium discovery company. Nuclear is the only clean, scalable, and baseload energy source available today, and we’re structurally short on uranium. We're addressing that by applying AI/ML to a discovery process that hasn't fundamentally changed in decades.
Backed by General Catalyst, 776 Ventures, Y Combinator, and others.
**The role**
You'll be building the ML systems at the core of Terranox. The work spans multiple modalities and model classes, much of it without an off-the-shelf playbook. You'll train computer vision models from scratch on messy, sparsely-labeled data and ship them into a production pipeline that drives real-world decisions. This is a hands-on, build-from-zero role.
**Requirements**
* 3+ years of ML experience, with a track record of building ML systems from scratch
* Strong computer vision background
* Experience taking ML systems to production
* Comfort framing ambiguous problems, not just solving well-defined ones
* Fluency with recent ML research and the ability to translate it into shipped systems
**Bonus but not required**
* Geospatial experience
* Background in sparse-label or weak-supervision regimes
* Prior early-stage startup experience
**Logistics**
* In-person in San Francisco