6 months ago
Base Salary
$230k - $322k/yr
Responsibilities
- Design end-to-end MLOps model lifecycle patterns covering data preparation, model management, and experiment tracking.
- Develop and support a graph ML codebase and platform that abstracts common patterns and improves model scalability and iteration.
- Collaborate with ML engineers to improve model training time, efficiency, and GPU training costs in a large distributed training environment.
- Optimize batch data processing in a data warehouse using Apache Beam, Apache Spark, Ray Data, and related tools.
- Architect pipelines that build and maintain graph data structures containing billions of nodes and tens of billions of edges.
- Lead development of a platform for large-scale machine learning models at Reddit.
Requirements
- At least 8 years of experience in machine learning infrastructure, including model training and deployments.
- Hands-on experience with machine learning optimization, memory profiling, and GPU profiling.
- Deep experience with cloud technologies for ML platforms, including GCP BigQuery, Google Cloud Storage, and Terraform.
- Hands-on experience administering and integrating MLOps tools for experiment tracking, model serving, and model registries such as MLflow or Wandb.
- Proficiency with Python, PyTorch, TensorFlow, and other common ML programming languages and frameworks.
- Deep experience with distributed training frameworks, including Ray and Kubernetes.
- Strong focus on scalability, reliability, performance, usability, and advocacy for platform users.
- Strong organizational and communication skills.
- Experience with graph databases such as Neo4j, JanusGraph, or TigerGraph is preferred.
- Experience with graph neural networks and graph ML frameworks such as PyTorch Geometric or Deep Graph Library is preferred.
Benefits
- U.S.-based employees receive medical, dental, and vision insurance.
- 401(k) program with employer match.
- Generous vacation time and parental leave.
- The role may be eligible for equity in the form of restricted stock units and, depending on the position offered, commission.
- Interviews may be recorded, transcribed, and summarized by artificial intelligence, with an option to opt out before scheduled interviews.
Tech Stack
Categories
About Reddit
Reddit builds a social news and discussion platform organized into user-created forums (“subreddits”) for consumers, creators, and communities. It monetizes primarily through advertising tools for brands and self-serve advertisers, plus a Premium subscription; developers can access data via its API. Founded in 2005 and headquartered in San Francisco, Reddit became a public company on the NYSE in 2024 and hosts over 100,000 active communities.