Principal Software Engineer, Inference

HPE's Private Cloud AI organization is seeking a Principal Software Engineer to lead the model runtime within HPE AI Essentials, the inference platform used by enterprises to operate large language models on infrastructure they own, including air-gapped and sovereign environments. The principal engineering challenge in this domain is not model deployment but sustained execution efficiency: achieving low tail latency and high GPU utilization on customer-owned hardware of varying generation and configuration. In this role you will define the architecture of that runtime – engine integration, batching, KV cache management, and distributed execution – together with the Kubernetes orchestration layer that supports it. The primary work location is as listed, but could be any other HPE site location in the US; however, remote work options will be considered.

·       Define and own the technical direction of the LLM serving deployment, including engine integration, continuous batching, KV cache management and reuse, and quantized execution

·       Partner with inference performance engineering teams, with accountability for time-to-first-token, inter-token latency, throughput per GPU, and P95/P99 tail latency

·       Define distributed inferencing strategy, including disaggregated prefill/decode, tensor and pipeline parallelism, KV cache offload across GPU memory, host memory, and RDMA-attached storage

·       Evaluate emerging runtimes, quantization schemes, speculative decoding, and mixture-of-experts serving, and determine whether each runtime is adopted, developed in-house, or declined

·       Define the orchestration layer supporting the runtime, including model admission, GPU scheduling and partitioning, cache-aware request routing, and autoscaling

·       Mentor engineers, lead design and architecture reviews, and present technical direction to business unit and executive audiences

·       Production experience with LLM inference engines such as vLLM, SGLang, TensorRT-LLM, TGI, or NVIDIA NIM, including modification of engine internals

·       Comprehensive understanding of inference internals, including continuous batching, paged attention, KV cache reuse and prefix caching, chunked prefill, quantization, and speculative decoding

·       Tensor and pipeline parallelism, NCCL collective operations, and the GPU memory hierarchy and interconnect characteristics that govern them

·       Expert level proficiency in Kubernetes platform architectures, including operators, custom resources, controllers, and scheduling

·       Strong programming proficiency in Go and Python, with the ability to read, debug, and profile C++/CUDA using tools such as Nsight

·       Experience with debugging/profiling multi-tier application workloads such as RAG, Agents, etc

·       Excellent analytical, debugging, and problem-solving abilities

·       Upstream contribution to vLLM, SGLang, TensorRT-LLM, LLM-D, LMCache, or KServe

·       Disaggregated prefill/decode serving, or KV cache offload and reuse at scale

·       RDMA, GPUDirect Storage, InfiniBand, or RoCE

·       MIG, fractional GPU allocation, and multi-tenant GPU isolation

·       On-premises, air-gapped, or regulated enterprise software delivery

·       Minimum of 12 years of experience in Software Engineering, including +1 years working directly on LLM inference runtimes or production model serving

·       Degree in Computer Science or related field

HPE is committed to creating an inclusive and accessible workplace and encourages applications from all qualified individuals, including those with disabilities. If you believe you require accommodation during any stage of the application or interview process, please submit your request by completing our secure form linked here.

Note: This option is reserved for applicants needing assistance/reasonable accommodation related to a disability.

We strive to provide our team members and their loved ones with a comprehensive suite of benefits that supports their physical, financial and emotional wellbeing.

We also invest in your career because the better you are, the better we all are. We have specific programs catered to helping you reach any career goals you have — whether you want to become a knowledge expert in your field or apply your skills to another division.

We are unconditionally inclusive in the way we work and celebrate individual uniqueness. We know varied backgrounds are valued and succeed here. We have the flexibility to manage our work and personal needs. We make bold moves, together, and are a force for good.

Information about employee benefits offered in the US can be found at https://myhperewards.com/main/new-hire-enrollment.html

HPE will comply with all applicable laws related to employer use of arrest and conviction records, including laws requiring employers to consider for employment qualified applicants with criminal histories.

We have become aware of an increase in fraudulent recruitment activities in which individuals impersonate our company or authorized recruitment agencies to offer fake employment opportunities. These scams may occur through false websites, emails, social media, or chat-based applications and often aim to obtain personal information or money. Please note that Hewlett Packard Enterprise (HPE), its direct and indirect subsidiaries and affiliated companies, and its authorized recruitment agencies/vendors will never charge a candidate a registration fee, hiring fee, or any other fee in connection with its recruitment and hiring process. We also never request personal information such as back account details, Social Security numbers, or national IDs via social media or chat applications.

All legitimate job opportunities will come through official company channels, and candidates are responsible for verifying the credentials of any third party claiming to represent the company. Any reliance on fraudulent communication is at the individual’s own risk, and HPE disclaims legal liability for any resulting damages. If you suspect recruitment fraud, do not share personal information or make any payments and report the incident to your local authorities immediately.