Machine Learning Engineer; Physics-AI

Qualifications

• We are looking for a Physics-AI-oriented Machine Learning Engineer who can do some Computer Vision, rather than a Computer Vision-focused MLE that can do some Physics-AI
• On the Applied Science team, we build machine learning models that power Stand’s risk analytics and climate resilience platform
• Strong foundation in ML: proven track record of taking models from research to production (training from scratch, fine-tuning advanced architectures, evaluation across diverse datasets)
• Hands-on experience combining physics-based modeling (finite element, finite volume, finite difference)
• with machine learning to accelerate or enhance solvers
• Expertise with computer vision and multimodal architectures (e.g., CNNs, ViTs, spatial attention, GNNs)
• Familiarity with modern generative and 3D methods (e.g., diffusion, autoregressive, 3D reconstruction)
• Proficiency in ML frameworks (
• containers, CI/CD, automated testing, shared libraries
• Ability to stay current with emerging methods and proac…

Benefits

• $160K – $210K

Responsibilities

• As a Machine Learning Engineer, you’ll own projects end-to-end — designing, training, and deploying models that deliver both immediate and long-term business impact
• You’ll thrive in a fast-moving startup environment, collaborating across Applied Science and the broader company to turn technical breakthroughs into production-ready solutions
• Your focus will be on creating multimodal, physics-aware models that accelerate physical modeling by orders of magnitude, adapt to diverse perils, and leverage the latest methods in the field
• If you’re eager to push the boundaries of applied machine learning, contribute to scaling classical simulation methods by 1000x, and help provide insurance for homes in climate-stressed areas—all while creating immense value from the ground up—this position is for you!
• You’ll partner with other MLEs and drive forward initiatives such as:
• Developing flagship physics-informed deep learning models
• Advancing multimodal modeling with data augmentation and sensor fusion
• Applying 3D computer vision for digital twin annotation
• Scaling spatial data analysis into production workflows
• This Role Will
• Design, train, and deploy ML models across physics-informed AI, computer vision, and multimodal learning
• Own projects end-to-end
• , from prototyping through production
• , with emphasis on reliability and business-ready tooling
• Fine-tune and extend state-of-the-art models to accelerate simulation and digital twin pipelines
• Build scalable ML infrastructure
• : data pipelines, training methodologies, and evaluation frameworks for real-time risk analytics
• Collaborate cross-functionally to integrate models into user-facing workflows
• Continuously improve model performance through monitoring, retraining, and active learning
• ) and production-grade practices (

Full Description

Position: Machine Learning Engineer (Physics-AI)

Location

San Francisco

Employment Type

Full time

Location Type

On-site

Department

Science & Engineering

Compensation

• $160K – $210K

• Offers Equity

About Stand

Stand is a new technology and insurance company revolutionizing how society assesses, mitigates, and adapts to climate risks.

Our leadership team has extensive experience in insurance,technology, and climate science: building billions in market value at prior ventures.

At Stand, we are rethinking how insurance enables proactive, science-driven resilience.

Existing insurance models often rely on broad exclusions, leaving homeowners without options.

At Stand, we leverage advanced deterministic models and cutting-edge analytics to provide personalized risk assessments—helping homeowners secure coverage and take proactive steps toward resilience.

Background

Homes respond differently to climate catastrophes like wildfire — but until now, we’ve lacked the tools to measure that risk at the individual level.

At Stand, we combine deterministic physics models with cutting-edge AI to deeply understand a home’s unique risk environment.

This enables broader insurance access, incentivizes proactive mitigation, and helps communities become more resilient.

The Role

We are looking for a Physics-AI-oriented Machine Learning Engineer who can do some Computer Vision, rather than a Computer Vision-focused MLE that can do some Physics-AI

.

On the Applied Science team, we build machine learning models that power Stand’s risk analytics and climate resilience platform.

We combine AI, embedded physics, and spatial intelligence into scalable tools that directly influence underwriting, pricing, and customer decision-making.

As a Machine Learning Engineer, you’ll own projects end-to-end — designing, training, and deploying models that deliver both immediate and long-term business impact.

You’ll thrive in a fast-moving startup environment, collaborating across Applied Science and the broader company to turn technical breakthroughs into production-ready solutions.

Your focus will be on creating multimodal, physics-aware models that accelerate physical modeling by orders of magnitude, adapt to diverse perils, and leverage the latest methods in the field.

If you’re eager to push the boundaries of applied machine learning, contribute to scaling classical simulation methods by 1000x, and help provide insurance for homes in climate-stressed areas—all while creating immense value from the ground up—this position is for you!

You’ll partner with other MLEs and drive forward initiatives such as:

• Developing flagship physics-informed deep learning models

• Advancing multimodal modeling with data augmentation and sensor fusion

• Applying 3D computer vision for digital twin annotation

• Scaling spatial data analysis into production workflows

This Role Will

• Design, train, and deploy ML models across physics-informed AI, computer vision, and multimodal learning

• Own projects end-to-end

, from prototyping through production

, with emphasis on reliability and business-ready tooling

• Fine-tune and extend state-of-the-art models to accelerate simulation and digital twin pipelines

• Build scalable ML infrastructure

: data pipelines, training methodologies, and evaluation frameworks for real-time risk analytics

• Collaborate cross-functionally to integrate models into user-facing workflows

• Continuously improve model performance through monitoring, retraining, and active learning

Core Skills (Must-Haves)

• Strong foundation in ML: proven track record of taking models from research to production (training from scratch, fine-tuning advanced architectures, evaluation across diverse datasets)

• Hands-on experience combining physics-based modeling (finite element, finite volume, finite difference)

with machine learning to accelerate or enhance solvers

• Expertise with computer vision and multimodal architectures (e.g., CNNs, ViTs, spatial attention, GNNs)

• Familiarity with modern generative and 3D methods (e.g., diffusion, autoregressive, 3D reconstruction)

• Proficiency in ML frameworks (

PyTorch/Tensor Flow

) and production-grade practices (

containers, CI/CD, automated testing, shared libraries

)

• Ability to stay current with emerging methods and proac…