About this role
About Ghost
Ghost is a personal AI hardware company.
To realize the full benefit of what models will do for people, we believe an entirely new platform will be built around AI. Some devices will let your AI understand what is happening around you, others will enable you to get things done without navigating apps or menus.
As these systems become more capable and more personal, it matters who owns them. We’re building a future where that intelligence belongs to you, with an ecosystem of devices and software that gives you greater control over your time, attention, and the things you want to accomplish.
What you'll do
Build and optimize the inference stack that runs personal AI models locally on Core.
Improve latency, throughput, and memory efficiency across interactive requests and background agent workloads.
Profile performance across the hardware and software stack, optimizing kernels, memory movement, and runtime execution.
Build benchmarks that reflect real product workloads, including long contexts, concurrent agents, and sustained operation.
Work closely with the hardware, post-training, and product teams.
Who you are
Experience building or optimizing inference systems for language models.
Strong systems programming skills in C++ or similar languages, alongside proficiency in Python.
Familiarity with GPU architecture, memory management, and performance profiling.
Experience with inference runtimes, quantization, or GPU kernel development using CUDA, Triton, or equivalent tools.
Strong experimental judgment: you can identify bottlenecks and evaluate tradeoffs between speed, memory, power, and model quality.
Ability to work on-site in San Francisco.