Note: The job is a remote job and is open to candidates in USA. AMD builds computing products and software that accelerate next-generation experiences across AI, data centers, PCs, gaming, and embedded systems. The Senior Product Manager will own the ROCm inference frameworks product strategy and roadmap, translating customer, ecosystem, and engineering needs into software priorities for production AI inference on AMD Instinct hardware. The role partners with engineering teams, open-source communities, customers, and ecosystem partners to drive scalable, performant, and reliable inference solutions.
Responsibilities
- Own the product strategy and roadmap for ROCm’s inference frameworks software capabilities
- Define how AMD’s framework-layer software stack for inference enables production workloads from single-GPU through rack-scale deployments, with a focus on performance, developer experience, portability, and competitive differentiation
- Identify emerging model architectures, serving technologies, and infrastructure shifts that require changes in the inference frameworks layer and translate them into product priorities
- Balance customer needs, ecosystem direction, technical opportunities, and engineering investment to determine where AMD should differentiate in the software stack
- Partner with engineering teams across inference engines, serving and orchestration, memory systems, GPU libraries, runtimes, kernels, and communications, with particular emphasis on the framework and orchestration layers that enable large-scale inference above the underlying GPU software stack
- Translate customer and ecosystem needs into prioritized engineering requirements and drive execution across organizational boundaries
- Work with engineering to identify inference performance bottlenecks and determine where framework-level software investment can have the greatest impact
- Provide software-informed input to future GPU architecture and platform decisions relevant to inference performance at scale
- Drive release readiness across a fast-moving open-source software ecosystem, with attention to performance, compatibility, regression risk, and adoption
- Serve as an active AMD presence in the open-source inference community
- Build relationships with maintainers, developers, researchers, and contributors whose work influences AI infrastructure and AMD adoption
- Track emerging technologies and ecosystem direction through open-source projects, research, technical communities, and developer feedback
- Communicate AMD’s roadmap and technical direction through community engagement, technical writing, and industry events
- Represent AMD constructively in technical discussions where ecosystem and company priorities may differ
- Work directly with sophisticated AI infrastructure customers to understand deployment requirements, performance constraints, and future needs at the framework and serving layer
- Distinguish individual customer requests from broader technical and market signals, then translate those signals into scalable product decisions
- Balance the needs of strategic customers with those of the broader open-source ecosystem
- Collaborate with cloud providers, model developers, infrastructure companies, and other ecosystem partners on joint integrations and solutions
Skills
- Deep understanding of production AI inference systems, including inference engines such as vLLM, SGLang, and AMD ATOM, and serving and orchestration systems such as llm-d, along with memory management and distributed inference
- Experience driving complex technical initiatives across multiple engineering teams or organizations
- Experience working with sophisticated AI infrastructure customers, open-source communities, or ecosystem partners
- Experience in AI inference infrastructure, GPU computing, distributed systems, cloud infrastructure, high-performance computing, developer platforms, research, or technical product leadership
- Ability to communicate technical tradeoffs clearly to engineering leaders, customer stakeholders, open-source contributors, and executive audiences
- Candidates from engineering, research, or other technical leadership backgrounds are encouraged if they demonstrate strong product instincts and interest in product leadership
- Bachelor's degree in Computer Science, Electrical Engineering, or a related technical field
- This role is not eligible for visa sponsorship
- Open-source contributions, technical writing, conference talks, or sustained community engagement are strong positive signals
- Familiarity with AMD Instinct™ hardware and the ROCm software ecosystem is a plus
- Product management experience in deeply technical products is valuable but not required
- Advanced degree a plus but not required given equivalent experience
Benefits
- Remote work arrangement
Company Overview