Senior ML Infra Engineer — Train & Deploy Vision Models

Voxel Sa; San Francisco, Northern (CA, KY); Hybrid; USD 140,000 - 210,000; Full time

Benefits offered by this job

Equity via equity plan

Health benefits

Unlimited PTO

Daily meals in-office

Parental leave

Job summary

Voxel, based in San Francisco, seeks a strong software engineer to own the ML infrastructure that powers how we train and ship vision models. You’ll build scalable training systems, manage experiments, and export models to optimized formats like TensorRT/ONNX, collaborating with CV, ML Data, and Platform engineers.

You will establish ML experiment tracking, embrace DevOps-for-ML on AWS, and design scalable solutions that accelerate research-to-production.

Qualifications

• 4+ years of experience building and shipping large scale software solutions.
• Hands-on experience building ML training pipelines in PyTorch.
• Experience with ML experiment tracking tools (Weights & Biases, MLflow, ClearML).
• Experience with AWS (S3, EC2, EKS) for ML workloads.
• Strong Python; write production-grade code.
• Track record of owning infrastructure end-to-end.
• Bias toward shipping and practical results.

Responsibilities

• Build and maintain ML training infrastructure to train multiple models concurrently and manage experiments.
• Own the train-to-deploy handoff—export trained models to optimized formats and quantify impact for production deployment.
• Establish ML experiment tracking and lifecycle management with tools like Weights & Biases, MLflow, or ClearML.
• Implement DevOps-for-ML best practices on AWS (IaC, CI/CD, observability, cost monitoring).
• Design scalable infra solutions that support model development for CV teams.

Skills

ML training pipelines

Python development

DevOps practices

Communication

End-to-end ownership

Tools

PyTorch

Weights & Biases

MLflow

ClearML

AWS

Job description

Voxel, based in San Francisco, seeks a strong software engineer to own the ML infrastructure that powers how we train and ship vision models. You’ll build scalable training systems, manage experiments, and export models to optimized formats like TensorRT/ONNX, collaborating with CV, ML Data, and Platform engineers.

You will establish ML experiment tracking, embrace DevOps-for-ML on AWS, and design scalable solutions that accelerate research-to-production.