Application Software Engineer, Inference

SpaceX was founded under the belief that a future where humanity is out exploring the stars is fundamentally more exciting than one where we are not. Today SpaceX is actively developing the technologies to make this possible, with the ultimate goal of enabling human life on Mars. APPLICATION SOFTWARE ENGINEER, INFERENCE The application software team is the central nervous system of SpaceX – we create mission critical applications that are used throughout SpaceX to accelerate launch vehicle production and flight as well as systems that allow Starlink to grow into a worldwide fast, reliable Internet service. We are looking for engineers who treat fellow teammates with fairness, respect, and support. Our team maintains a high-performance AI inference platform that serves the best models internally at SpaceX to accelerate our most ambitious engineering goals. As part of this effort in Palo Alto, you will design and optimize large-scale model serving systems end-to-end, owning everything from distributed infrastructure to deep low-level optimizations. You will work on systems that deliver reliable, high-throughput inference to power SpaceX’s mission-critical applications while maintaining the highest standards of performance and availability. Aerospace experience is not required to be successful here - rather we look for smart, motivated, respectful, collaborative engineers who love solving problems and want to make an impact on a super inspiring mission. You will have full ownership of challenging problems, working with a team of enthusiastic engineers with diverse perspectives to design and produce solutions that enable SpaceX to achieve its loftiest engineering goals at a rapid pace. The success of the missions at SpaceX depends on the software that you and your team produce. This role will report through SpaceX Application Software while also working closely with xAI engineering teams. RESPONSIBILITIES: Develop highly reliable, high-throughput inference systems that serve the best AI models internally across SpaceX Architect and implement scalable distributed infrastructure for model serving, including load balancing, auto-scaling, batch scheduling, global KV cache, and continuous batching Optimize latency and throughput of model inference under real production workloads, including low-level GPU kernel work, quantization, speculative decoding, and other acceleration techniques Build reliable, high-concurrency serving systems with 100% uptime, low tail latency, and excellent observability Own end-to-end components such as request routing, SDK development, rate limiting, and efficient scaling for internal SpaceX AI inference platforms Benchmark, fine-tune, and accelerate inference engines (e.g., SGLang, vLLM, TensorRT-LLM) Develop custom tools for tracing, replaying, and resolving issues across the full stack — from orchestration down to GPU kernels Create robust CI/CD infrastructure for seamless endpoint deployment, image publishing, and inference engine updates Collaborate across SpaceXAI teams to integrate inference capabilities into broader systems and workflows BASIC QUALIFICATIONS: Bachelor's degree in computer science, engineering, math, or scientific discipline; OR 2+ years of professional experience building software in lieu of a degree Experience in designing, implementing, and maintaining reliable and horizontally scalable distributed systems 1+ years of experience in full stack development or backend development with production systems 1+ years of experience with Rust or C++ PREFERRED SKILLS AND EXPERIENCE: Experience with LLM inference engines and serving frameworks (e.g., SGLang, vLLM, Triton, TensorRT-LLM) Deep low-level systems programming and optimizations: GPU kernels, code generation, batching, caching, parallelism, quantization, and speculative decoding Experience with large-scale, high-concurrency production serving systems Knowledge of service observability and reliability best practices Experience operatin…