Software Engineer - Training/Inference (C++)

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 making life multiplanetary and expanding consciousness to the stars. SpaceXSI is building Super Intelligence to help humanity understand the universe. Our teams are building the full Super Intelligence stack: from our Colossus supercomputers, to Grok models, to our suite of applications, including Grok Bot, Grok Build, Grok Imagine, Grok Voice, and X. We're creating a powerful flywheel where more compute trains better models, better models power more useful products, and what we learn from people using those products shapes what we build next. ABOUT THE ROLE: We are building one of the world’s largest AI supercomputers from the ground up. As part of the Compute Infrastructure team, you will own both the raw GPU supercomputer and the platform layer that runs on top of it. You will work across the full stack — from low-level GPU kernel optimizations and Linux kernel internals to massive-scale orchestration and virtualization — to make training and inference at SpaceXSI as fast, reliable, and scalable as possible. This is a broad, high-impact role that combines hardcore supercompute and compute infrastructure work. Your contributions will directly accelerate Grok’s training speed and overall AI progress. RESPONSIBILITIES : Design, build, and optimize massive GPU clusters for extreme-scale training and inference workloads Develop and tune low-level CUDA kernels (GeMM, Attention, etc.), using CUTLASS, Tensor Cores, and Nsight for maximum performance Profile, debug, and eliminate bottlenecks across GPU memory hierarchy, networking fabric, filesystems, and multi-GPU operation Collaborate closely with AI research teams to deliver production-grade performance and scalability PREFERRED SKILLS AND EXPERIENCE: Deep low-level syst

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 making life multiplanetary and expanding consciousness to the stars. SpaceXSI is building Super Intelligence to help humanity understand the universe. Our teams are building the full Super Intelligence stack: from our Colossus supercomputers, to Grok models, to our suite of applications, including Grok Bot, Grok Build, Grok Imagine, Grok Voice, and X. We're creating a powerful flywheel where more compute trains better models, better models power more useful products, and what we learn from people using those products shapes what we build next. ABOUT THE ROLE: We’re looking for exceptional storage & database engineers who want to join us on a new project to deeply integrate SpaceXSI's advanced AI infrastructure into a platform used by around 600 million users every month. We’re building a new storage tier that powers training, inference, recommendations, and real-time data extraction. We build and operate: An exabyte-scale S3-compatible object store A multi-tenant transactional key/value store that serves 100s of millions requests per second A caching tier that serves billions of requests per second A scalable vector database that powers our recommendation systems A few other specialized databases serving hundreds of millions requests per second RESPONSIBILITIES: Design, build, and launch to production new features and improvements aimed at unifying common components across the storage systems Dive into performance issues and work with customers and deliver solutions to cater to customers’ latency, availability and data durability requirements. Lead and drive incident responses and recovery with your peers. Review and contribute to incident postmortems and hold a high bar for the same. Work

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 making life multiplanetary and expanding consciousness to the stars. SpaceXSI is building Super Intelligence to help humanity understand the universe. Our teams are building the full Super Intelligence stack: from our Colossus supercomputers, to Grok models, to our suite of applications, including Grok Bot, Grok Build, Grok Imagine, Grok Voice, and X. We're creating a powerful flywheel where more compute trains better models, better models power more useful products, and what we learn from people using those products shapes what we build next. ABOUT THE ROLE: As part of the Network Software and Services for AI (nssAI) team at SpaceXSI , you'll build cutting-edge software, services, and frameworks to empower our Network Development Engineers. Working hands-on, you’ll tackle all facets of network management—metric collection, configuration, zero-touch provisioning, monitoring, and auto-remediation—driving automation-first solutions for SpaceXSI's production and ancillary networks. Expect to develop extensible tools, streamline complex processes, and ensure rock-solid reliability to support SpaceXSI's mission of accelerating human scientific discovery through AI. RESPONSIBILITIES: Building software and tools with extensive metrics coverage for some of the world’s largest GPU supercomputing network fabrics used for AI training and serving customer inference queries. Implement IaC best practices, enhancing deployment pipelines, and ensuring robust, secure service delivery across our production environments. BASIC QUALIFICATIONS: Deep experience collaborating with network engineers daily using extensive knowledge of network topologies, physical and logical, and network protocols. Expert knowledge and proven history with des

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 making life multiplanetary and expanding consciousness to the stars. SpaceXSI is building Super Intelligence to help humanity understand the universe. Our teams are building the full Super Intelligence stack: from our Colossus supercomputers, to Grok models, to our suite of applications, including Grok Bot, Grok Build, Grok Imagine, Grok Voice, and X. We're creating a powerful flywheel where more compute trains better models, better models power more useful products, and what we learn from people using those products shapes what we build next. ABOUT THE ROLE: SpaceXSI is building and operating large-scale networks that underpin training and inference infrastructure, including high-performance / supercompute fabrics that connect GPU clusters, plus the core, edge, and datacenter networks that keep that infrastructure reachable and reliable. We need a Network Engineer who is strong on fundamentals and comfortable owning production network design, deployment, and operations end to end. This is a hands-on engineering seat — not a NOC technician role and not a network-software (telemetry/ZTP platform) SWE role. You will design and build networks, qualify platforms, ship changes safely, and keep availability and performance high as we scale. Travel to Memphis (and other build sites) may be required for capacity build-outs. You will participate in a team on-call rotation. RESPONSIBILITIES: Design, deploy, and operate production datacenter and campus/core networks at scale Own routing and switching configuration standards (BGP and at least one IGP such as OSPF or IS-IS), including change design, peer reviews, and execution Qualify new network platforms, optics, and topologies; contribute to architecture and capacity planning &

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 making life multiplanetary and expanding consciousness to the stars. SpaceXSI is building Super Intelligence to help humanity understand the universe. Our teams are building the full Super Intelligence stack: from our Colossus supercomputers, to Grok models, to our suite of applications, including Grok Bot, Grok Build, Grok Imagine, Grok Voice, and X. We're creating a powerful flywheel where more compute trains better models, better models power more useful products, and what we learn from people using those products shapes what we build next. You will own and elevate our mission-critical research platform that directly accelerates SpaceXSI's frontier work. The evals team creates and executes instruments for measuring model capabilities and behaviors. These instruments are used to clean datasets, provide feedback for experiments, and track growth throughout training. The evals team also diagnoses why agents fail and how to make improvements both short and long term. We manage the execution infrastructure, design data sets and grading schemes. You will design, build, and ship delightful and bulletproof systems that engineers love to use every day. This role values a strong attention to details, first principles reasoning, and great taste. We are open to all levels from new grads to senior engineers. Requirements Software engineering proficiency and 3+ years of experience actively writing code. A track record of shipping high quality projects and a strong sense of urgency. Exceptional product taste - you have a sharp sense for what “good” looks like in outputs, and you can translate fuzzy notions of quality, helpfulness, truthfulness, and safety into concrete, measurable criteria. Strong attention to d

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 making life multiplanetary and expanding consciousness to the stars. SpaceXSI is building Super Intelligence to help humanity understand the universe. Our teams are building the full Super Intelligence stack: from our Colossus supercomputers, to Grok models, to our suite of applications, including Grok Bot, Grok Build, Grok Imagine, Grok Voice, and X. We're creating a powerful flywheel where more compute trains better models, better models power more useful products, and what we learn from people using those products shapes what we build next. ABOUT THE ROLE: At SpaceXSI, we design, build, and operate Colossus from the ground up. This includes the massive GPU clusters, high-speed interconnect fabric, and the software that makes it all work at unprecedented scale. Colossus powers Grok and our frontier AI models with a custom, high-performance datacenter network that delivers ultra-low latency and massive bandwidth across hundreds of thousands of GPUs. As a Software Engineer on the Colossus Networking team, you will develop the core networking software that maximizes the performance and reliability of our datacenter fabric. Your work will directly impact training efficiency, model convergence, and the speed at which we can push the frontier of AI. Our engineers own the full lifecycle of their software — from design and implementation to deployment, monitoring, and iteration based on real-world performance at scale. You will solve hard problems in distributed systems, high-performance networking, and real-time control of one of the largest AI supercomputers on Earth. RESPONSIBILITIES: Develop routing and traffic-engineering algorithms for the Colossus high-performance datacenter network. Develop highly reliable, real-tim