SambaNova Systems

Launch your career on SambaNova’s Cloud Platform team by building the automation, infrastructure, and observability systems that keep AI inference services fast and available worldwide. This new-grad role offers hands-on experience with Kubernetes, Terraform, monitoring, regional expansion, and incident response.

Build production ML inference infrastructure for SambaNova’s RDU accelerators as a new-grad engineer. You’ll develop serving features, adapt vLLM, maintain public APIs, and improve accuracy, latency, throughput, and reliability.

Join SambaNova’s Runtime team as a new-grad engineer building the lowest-level software stack for advanced AI hardware. You’ll develop drivers, libraries, networking, and system tools in C/C++/Rust and Python for large-scale ML training and inference systems.

Build the compiler technology that powers SambaNova’s RDU AI accelerator, working across IR infrastructure, model lowering, tensor tiling, memory management, and operation mapping. This new-grad role offers hands-on work in C++ and Python alongside runtime and hardware teams.

Cloud Platform Architect responsible for building and scaling SambaNova’s internal developer platform across multi-cloud, on-premise, and Kubernetes environments. The role focuses on infrastructure automation, AI workload platforms, networking, observability, and production reliability.

Lead the reliability and scalability of SambaNova’s production AI inferencing platform across global cloud and on-premise environments. This senior role owns monitoring, incident response, capacity planning, infrastructure automation, and performance optimization for customer-facing inference services.

Build and operate production-grade ML inference infrastructure for SambaNova’s RDU accelerators, powering high-throughput, low-latency services through SambaStack and SambaCloud. This senior individual contributor role focuses on advanced decoding, caching, serving frameworks, APIs, and accuracy verification.

Join SambaNova as a New Grad AI Performance Engineer, bringing up and optimizing foundation models for high-throughput, low-latency inference on its dataflow platform. You’ll profile execution across software and hardware layers and apply techniques such as quantization, kernel fusion, and scheduling optimization.