AI Product Manager (Agents & Local LLMs)

Beijing / Shanghai / 3+ years in product management / Bachelor’s / Negotiable / 1 opening

Responsibilities

  1. Own application products for desktop AI computing devices, defining core use cases and the product roadmap from the ground up.
  2. Research how individuals, developers and small teams use AI, identifying where local computing delivers real value through privacy, offline access, persistent operation and cost.
  3. Track AI PCs, all-in-one devices, inference hardware and AI-native products; analyze competitors and assess product opportunities.
  4. Design agent products covering tool calling, task orchestration, knowledge retrieval, context and memory, long-running tasks, human intervention and error recovery.
  5. Plan the device software experience, including initial setup, model and application management, task execution, resource status, data integration and OTA updates.
  6. Coordinate software and hardware capabilities: define supported models, performance, concurrency and resource constraints based on the chip, inference runtime and complete system.
  7. Produce requirements, prototypes and acceptance criteria, working with software, algorithm, chip and design teams to deliver. Evaluate real tasks and track success rates, latency and retention.
  8. Collaborate with model and application partners on integration and joint development, exploring preinstalled apps, distribution and commercialization.

Requirements

  1. Bachelor’s degree or higher and at least three years in product management, with end-to-end delivery experience in AI applications, developer tools, desktop software or products combining software and hardware.
  2. Understand the capabilities and limitations of LLMs and agents. Be familiar with RAG, tool calling, context management, memory and evaluation, and able to discuss technical trade-offs with engineers.
  3. Understand the value of local deployment and clearly explain the relationship between devices, models and applications.
  4. Strong user research, competitive analysis and prioritization skills; able to translate technological changes into concrete use cases and validation plans.
  5. Able to explore new use cases from scratch, rapidly prototype, validate with real users and adjust investment based on the findings.
  6. Able to manage complex software experiences involving model deployment, task scheduling, file and data integration, permissions, and multiple users or devices.
  7. Understand how inference speed, GPU and system memory, context length, concurrency and power consumption affect the user experience, and set measurable targets.
  8. An avid AI product user with strong practical skills, able to read English documentation and independently evaluate new models and tools.

Nice to have

  1. Experience with AI PCs, local AI appliances, workstations or edge computing products.
  2. Experience deploying models locally with Ollama, llama.cpp or vLLM, building agents, packaging and distributing applications, or developing developer platforms.
  3. Able to independently build working prototypes using AI coding tools, Python or APIs.
  4. Experience with long-running agents, multimodal creation, private knowledge bases or shared AI services for teams.
  5. Experience collaborating with chip vendors, device manufacturers, model communities or open-source communities.
  6. A track record of innovation pilots, validating willingness to pay or bringing products to market.