Tasks:
- * Build agent memory and context management
- * Build AI platform products from 0 to 1
- * Build service architectures and APIs
- * Contribute to CUDAGen research and papers
- * Create domain knowledge bases and strategy templates
- * Design heterogeneous compute scheduling
- * Develop agent evaluation loops
- * Develop cloud-edge collaboration solutions
- * Develop LLM agent systems
- * Explore on-device agent runtimes
- * Implement agent orchestration and tool calling
- * Implement task scheduling and load balancing
- * Improve platform observability
- * Integrate AI platforms with chip software stacks
Perks/Benefits:
- + Cross-stack technical exposure
- + Real world product impact
- + Support for academic publications
Skills/Tech stack required:
[Agent evaluation] [Agent Frameworks] [Agent Orchestration] [Agent systems] [API Design] [Autogen] [CPU GPU] [CPU GPU NPU Scheduling] [CUDA] [Data Visualization] [Data Warehousing] [Device Inference] [Dify] [Docker] [Edge AI] [FastAPI] [Git] [GPU Computing] [GPU/NPU scheduling] [Heterogeneous computing] [Langchain] [Langgraph] [LLM Agents] [Load Balancing] [Multi-Agent] [Multi-Agent Systems] [NPU Scheduling] [Observability] [On-device Inference] [Prompt engineering] [Python] [Service architecture] [SQL Optimization] [Task Scheduling] [Tool-Calling]
Educational requirements:
[Bachelor's Degree]
Role(s):
[Agent Engineer] [AI/Agent Engineer] [AI Platform Engineer] [Engineer] [LLM Engineer] [Platform Engineer]