Staff ML Platform Engineer: Scale Graph ML & MLOps

Reddit, Inc.; San Francisco (CA); On-site; USD 230,000 - 322,000; Full time

Benefits offered by this job

Equity in RSUs

Medical, dental, and vision insurance

401(k) employer match

Generous time off and parental leave

Job summary

Reddit, Inc. is seeking a Staff ML Infrastructure Engineer to lead the development of a platform for large-scale ML models. You will design end-to-end MLOps patterns and contribute to a graph ML codebase and platform that enables greater model scalability and iteration.

You will collaborate on performance tuning, data processing optimization, and building pipelines for billions of graph nodes, shaping a scalable and reliable ML infrastructure.

Qualifications

• 8+ years of experience in ML infrastructure, including model training and deployments.
• Hands-on experience with ML optimization, including memory and GPU profiling.
• Deep experience with cloud-based technologies for ML platforms (GCP, storage, IaC).
• Experience with MLOps tools for experiment tracking, model serving, and registries.
• Proficiency with Python, PyTorch, TensorFlow, and distributed training frameworks (Ray/Kubernetes).
• Strong focus on scalability, reliability, and user-centric platform design.

Responsibilities

• Design end-to-end ML platform patterns (MLOps) to boost development velocity.
• Develop and support a graph ML codebase and platform for scalability.
• Collaborate on performance tuning, including GPU training costs in a distributed setup.
• Optimize batch data processing using Beam, Spark, Ray Data, etc.
• Architect pipelines for massive graph data structures with billions of nodes.

Skills

ML infrastructure

Model training

GPU profiling

Cloud platforms

Python

PyTorch

TensorFlow

Ray

Kubernetes

Distributed training

Communication skills

Graph databases

Graph ML tools

Tools

Terraform

MLflow

Wandb

Neo4j

JanusGraph

TigerGraph

PyTorch Geometric

Deep Graph Library

Job description

Reddit, Inc. is seeking a Staff ML Infrastructure Engineer to lead the development of a platform for large-scale ML models. You will design end-to-end MLOps patterns and contribute to a graph ML codebase and platform that enables greater model scalability and iteration.

You will collaborate on performance tuning, data processing optimization, and building pipelines for billions of graph nodes, shaping a scalable and reliable ML infrastructure.