Role overview

UniversalAGI is building foundation models that learn physics directly from data to transform engineering simulation. The company works on problems in computational fluid dynamics and reservoir engineering for major industrial and defense organizations. As an ML Engineer, you will own the execution layer of machine learning development, connecting research breakthroughs to production deployments.

Key responsibilities

Build and maintain data preprocessing and data generation pipelines that support model training and evaluation.
Run training and fine-tuning workflows end-to-end and iterate rapidly on performance improvements.
Design and execute benchmarking and evaluation suites to measure progress and quantify customer outcomes.
Work with PhD researchers to transform model architectures into repeatable, production-grade workflows.
Communicate results through metrics, dashboards, and documentation while maintaining reproducible, high-quality work.

Requirements

Strong software engineering fundamentals including clean code practices, debugging, reliability, and reproducibility.
Solid understanding of machine learning foundations with hands-on experience across the full ML lifecycle from data preparation through training, fine-tuning, and evaluation.
Demonstrated experience training or fine-tuning models in any domain, whether large language models, computer vision, physics-based models, or surrogate models.
High personal standards and measurable focus on continuous improvement in your deliverables.
Resourcefulness in balancing quick correctness against robust solutions, with clear tradeoff reasoning.
Comfort owning work end-to-end and maintaining accountability for measurable outcomes.

What makes this role worth considering

You will work alongside a compact team of 25 researchers and engineers in San Francisco, with direct collaboration with the founder and CEO and influence over company strategy. The role offers the chance to define physics AI at a critical inflection point, working on cutting-edge problems at the intersection of deep learning and physics simulation. The company operates an in-person-first culture with five days a week onsite, comprehensive benefits including daily office meals, commute and wellness stipends, health coverage, and significant equity alongside competitive salary.