ML Ops Engineer: Scalable AI Data Pipelines

Vinci; Palo Alto (CA); On-site; USD 170,000 - 230,000; Full time

Job summary

Vinci is building a scalable ML platform for physics-inspired simulations that power real hardware. You will own the pipeline from recipe to versioned, validated datasets used in training, extending GPU scheduling and multi-tenant execution.

This hands-on role emphasizes reliability and traceability in a production codebase. You will collaborate with infrastructure and research teams to optimize costs, ensure data quality, and accelerate campaigns across millions of simulations and training

Qualifications

• 5+ years building production data or ML pipelines, incl. 2+ years with GPU-scale workloads.
• Experience running large batch or simulation workloads with many long-running jobs.
• Strong Python and software engineering fundamentals with trusted test suites.

Responsibilities

• Scale the data generation pipeline by orders of magnitude for millions of simulations.
• Own dataset lineage and reproducibility across generation campaigns.
• Build data quality checks, coverage, statistics for every sample.
• Run and evolve the job platform with Temporal, Ray, and Kubernetes-based workloads.
• Manage training launches, checkpointing, and model registry integration.
• Collaborate with infrastructure to ensure compute and data capacity meets demand.
• Make costs visible and identify stages to optimize GPU utilization.
• Keep the codebase healthy with tests, CI speed, and safe migrations.

Skills

Python

Software engineering

Distributed systems

Data pipelines

GPU workflows

Tools

Temporal

Ray

Airflow

Argo

Kubernetes Jobs

Terraform

Ansible

Job description

Vinci is building a scalable ML platform for physics-inspired simulations that power real hardware. You will own the pipeline from recipe to versioned, validated datasets used in training, extending GPU scheduling and multi-tenant execution.

This hands-on role emphasizes reliability and traceability in a production codebase. You will collaborate with infrastructure and research teams to optimize costs, ensure data quality, and accelerate campaigns across millions of simulations and training