Staff Machine Learning Engineer, Reddit

• BS, MS, or PhD in Computer Science or a related field
• 5+ years of software engineering experience
• Strong proficiency in Python
• Profiency in at least one systems language (Go, C++, Rust, or Java) preferred
• Experience building distributed systems at scale
• Experience with machine learning infrastructure, training systems, or model serving platforms
• Deep understanding of performance engineering and systems optimization
• Strong debugging and profiling skills

Desirable

• The ML Efficiency team builds the infrastructure, tooling, and optimization systems that enable machine learning engineers and researchers to train, evaluate, deploy, and operate models efficiently at scale.
• We focus on improving developer productivity, reducing infrastructure costs, increasing hardware utilization, and accelerating experimentation across the company’s ML ecosystem
• Design and build systems that improve the efficiency of ML training and inference workloads
• Develop tooling that helps ML engineers debug, profile, optimize, and monitor model performance
• Improve GPU and general resource utilization through scheduling, resource management, caching, and workload optimization
• Partner with ML researchers and product teams to identify bottlenecks and drive performance improvements
• Build benchmarking frameworks and performance dashboards for training and serving systems
• Optimize distributed training infrastructure, data pipelines, and model serving architectures
• Lead cross-functional initiatives that improve the productivity of Reddit ML engineers
• Drive technical strategy for ML platform scalability, reliability, and cost efficiency
• What Success Looks Like;
• ML engineers can move from idea to experiment faster
• Training and inference costs decrease, performance increases, while model quality is maintained or improved
• GPU utilization and cluster efficiency increase
• Platform reliability improves as ML workloads scale
• Teams spend less time managing infrastructure and more time building models
• Average recommendation model size increases