Location: Hybrid | Bellevue, WA (downtown)
multiple roles available
A well-funded, rapidly growing AI infrastructure company is building a next-generation cloud platform designed to power the full lifecycle of artificial intelligence. The organization is developing a comprehensive AI infrastructure, platform, and services portfolio that supports the full spectrum of AI workloads—including large-scale compute, model training, fine-tuning, inference, and emerging agentic AI applications.
Backed by significant long-term investment, the company combines the speed, ownership, and innovation of a startup with the stability and resources of an established parent organization. Engineering teams are intentionally lean, highly collaborative, and AI-native, leveraging modern tooling and automation to build infrastructure capable of supporting the industry's most demanding AI workloads.
We're seeking GPU Performance / Kernel Engineer to optimize the data plane powering large-scale AI workloads. This role focuses on improving GPU utilization, reducing latency, and maximizing throughput across training and inference environments by tuning kernels, identifying performance bottlenecks, and driving efficiency across the GPU fleet.
This is a high-impact engineering role focused on extracting maximum performance from large-scale GPU infrastructure. You'll work at the intersection of GPU architecture, AI workloads, systems performance, and low-level optimization.
As part of a highly technical infrastructure team, you'll analyze workload behavior, optimize performance-critical code paths, and develop the techniques and tooling required to operate AI systems efficiently at scale.
This opportunity is ideal for engineers who enjoy deep technical challenges involving GPU computing, kernel optimization, distributed AI workloads, and hardware/software performance.
Profile, analyze, and optimize GPU kernels to improve latency, throughput, and overall utilization.
Identify and eliminate data-plane bottlenecks impacting GPU performance across large-scale AI workloads.
Tune performance-critical workloads across training and inference environments.
Work closely with AI infrastructure, machine learning, and platform engineering teams to understand workload characteristics and optimize system behavior.
Develop benchmarking methodologies and performance measurement practices across GPU infrastructure.
Evaluate emerging GPU technologies, performance tools, and optimization techniques as hardware platforms evolve.
Contribute to engineering practices that improve GPU efficiency, scalability, and reliability across the fleet.
Strong experience with GPU kernel development and performance optimization using technologies such as CUDA, ROCm, or comparable GPU programming frameworks.
Demonstrated experience improving GPU utilization, reducing latency, or increasing throughput for production AI workloads.
Strong understanding of GPU architecture, memory hierarchy, parallel computing, and the data path from application layer to hardware execution.
Experience profiling and debugging performance issues in complex AI or distributed computing environments.
Ability to independently own technically complex problems and drive solutions in a fast-moving engineering environment.
Strong systems programming and performance engineering mindset.
Experience optimizing workloads across multiple GPU platforms, including NVIDIA and AMD architectures.
Experience with GPU compiler technologies, runtime optimization, or low-level systems performance.
Contributions to open-source GPU performance projects, compiler tooling, or AI systems optimization.
Background working with large-scale AI training, inference platforms, HPC environments, or cloud GPU infrastructure.
Familiarity with GPU profiling and optimization tools such as Nsight Systems, Nsight Compute, ROCm profiling tools, or similar technologies.
Competitive base pay for Bellevue market
Certain roles are eligible for additional rewards, including merit increases, annual bonus, and long term incentives. These awards are allocated based on individual performance
U.S. based employees have access to medical, dental, and vision insurance, a 401(k) plan and company match, employees also receive per calendar year, paid holidays.
Hybrid role based in the Bellevue, WA area.
Approximately three days per week in the office.
Candidates elsewhere in the U.S. who are open to relocation are encouraged to apply.
U.S. work authorization is required. Visa sponsorship is not currently available.
Optimize the performance layer behind one of the industry's most advanced AI infrastructure platforms.
Work directly on GPU efficiency, kernel optimization, and large-scale AI workload performance.
Solve some of the hardest challenges in AI systems engineering—maximizing utilization, minimizing latency, and scaling compute efficiently.
Join early enough to influence architecture, tooling, and performance engineering practices.
Collaborate with world-class engineers building the infrastructure powering the next generation of AI applications.
Enjoy the technical ownership and impact of a startup environment backed by significant long-term investment.
Build and operate a secure, reliable AI agents platform, owning its runtime, GenAI gateway, rollout controls, observability, and evaluation infrastructure for healthcare products.
Designs and implements generative AI solutions for customer environments, taking LLM, RAG, and agent-based systems from prototypes to scalable production integrations across cloud platforms and business processes.
Desenvolve e mantém aplicações mobile escaláveis e de alta performance com Flutter, integrando serviços de IA generativa, agentes e soluções RAG, além de contribuir para arquitetura, APIs e pipelines de CI/CD.
Lidera a arquitetura e evolução de uma plataforma de agentes de IA, desenvolvendo orquestração e recuperação de conhecimento e garantindo qualidade, segurança e confiabilidade em produção.
Desenvolve e evolui soluções de IA integradas ao OMS, usando LLMs, prompts e pipelines RAG para automatizar fluxos, otimizar o processamento de pedidos e reduzir trabalho manual operacional.
