ML Engineer

About the Role

Qualifications

• Strong software engineering skills (clean code, debugging, reliability, reproducibility).
• Solid ML foundations and hands-on experience with the ML lifecycle: data → training/fine-tuning → evaluation/benchmarking.
• Prior experience training or fine-tuning models (any modality/type - LLMs, computer vision, physics, surrogate models, etc.)
• Olympic athlete mindset: You have high standards for yourself and are obsessed with measurable improvement on the metrics you are delivering.
• Resourcefulness: you know when to do the “quick & correct” fix vs. when to invest in a robust solution, and you can justify the tradeoff with impact/
• Ownership: Comfortable owning work end-to-end and being accountable for measurable outcomes.
• Bonus Qualifications
• Experience building data pre-processing pipelines for training ML models.
• Experience with benchmarking methodology, experiment design, and metric selection.
• Familiarity with distributed training / scalable compute workflows.
• Experience in an FDE-style / delivery execution role (or similar “ship results fast” environments).
• Cultural Fit
• Technical Respect: Ability to earn respect through hands-on technical contribution
• Intensity: Thrives in our unusually intense culture - willing to grind when needed
• Customer Obsession: Passionate about solving real customer problems, not just publishing papers
• Deep Work: Values long, uninterrupted periods of focused work over meetings
• High Availability: Ready to be deeply involved whenever critical issues arise
• Communication: Can translate complex model decisions to customers and team
• Growth Mindset: Embraces the compounding returns of intelligence and continuous learning
• Startup Mindset: Comfortable with ambiguity, rapid change, and wearing multiple hats
• Work Ethic: Willing to put in the extra hours when needed to hit critical milestones
• Team Player: Collaborative approach with low ego and high accountability
• Bias for Action: Ships experiments fast, learns from failures, and iterates quickly
• What We Offer
• Opportunity to define the future of physics AI from the ground up
• Work on cutting-edge problems at the intersection of deep learning and physics simulation
• Direct collaboration with the founder & CEO and ability to influence company strategy
• Competitive compensation with significant equity upside
• In-person first culture - 5 days a week in office with a team that values face-to-face collaboration
• Access to world-class investors and advisors in the AI space

Benefits

• We provide great benefits, including:
• Competitive compensation and equity.
• Competitive health, dental, vision benefits paid by the company.
• 401(k) plan offering.
• Flexible vacation.
• Team Building & Fun Activities.
• Great scope, ownership and impact.
• AI tools stipend.
• Monthly commute stipend.
• Monthly wellness / fitness stipend.
• Daily office lunch & dinner covered by the company.
• Immigration support.
• How We’re Different
• “The credit belongs to the man who is actually in the arena, whose face is marred by dust and
• sweat and blood; who strives valiantly; who errs, who comes short again and again... who at the
• best knows in the end the triumph of high achievement, and who at the worst, if he fails, at least
• fails while daring greatly." - Teddy Roosevelt
• At our core, we believe in being “in the arena. ” We are builders, problem solvers, and risk-takers who show up every day ready to put in the work: to sweat, to struggle, and to push past our limits. We know that real progress comes with missteps, iteration, and resilience. We embrace that journey fully knowing that daring greatly is the only way to create something truly meaningful.
• If you're ready to train the models that will revolutionize physics simulation, push the boundaries of what AI can learn, and deliver real impact, UniversalAGI is the place for you.

Responsibilities

• Build and maintain data preprocessing and data generation pipelines to support model training and evaluation.
• Run training and fine-tuning workflows end-to-end and iterate quickly on performance improvements.
• Design and execute benchmarking/evaluation suites to measure progress and customer outcomes.
• Collaborate with PhD expert researchers to operationalize model architectures into repeatable, production-grade workflows.
• Communicate results clearly (metrics, dashboards, short writeups) and maintain high-quality, reproducible work.