Software Engineer, Distributed Systems

About the role

fal provides the infrastructure and orchestration layer for generative media applications. We are looking for an engineer to help us scale our platform to support high-traffic AI workloads and global low-latency requirements.

Key facts

What you'll do

• Develop core platform components using Python and Rust, including request routing, scheduling, GPU autoscaling, and queueing systems.
• Architect infrastructure capable of handling 100x traffic growth while maintaining low latency.
• Automate system maintenance and reliability tasks using AI-driven workflows.
• Optimize low-level CPU and memory performance through profiling and tuning.

Requirements

• Minimum 3 years of experience developing distributed compute and orchestration platforms using Python or Rust.
• Practical knowledge of distributed system principles such as consensus, fault tolerance, capacity planning, and scheduling.
• Proficiency in memory allocation and computational complexity.
• Proven history of deploying and managing systems at production scale.
• Experience utilizing observability tools to improve system reliability and performance.
• Ability to lead technical strategy and communicate effectively across the organization.
• Independent work ethic with a focus on rapid execution and ownership.

Nice to have

• Background in AI or ML inference and training infrastructure.
• Expertise in high-performance systems programming, including zero-copy, async runtimes, and memory-safe concurrency.
• Experience developing multi-tenant compute environments.
• Knowledge of networking performance and fundamentals.
• Familiarity with GPU scheduling constraints and workload characteristics.

Skills & tools

• Python
• Rust
• Distributed Systems
• GPU Orchestration
• Observability
• Systems Programming

Practical notes

• Benefits include health, dental, and vision insurance for US employees.
• Relocation assistance is available for those moving to San Francisco.
• The company hosts regular team events and offsites.