美商半導體大廠｜Agentic AI Architect｜年薪4.5M

WHAT YOU'LL DO

• Design and develop autonomous AI Agents capable of multi-step reasoning, planning, and tool execution to automate complex manufacturing workflows.
• Implement Agentic frameworks to orchestrate LLM interactions with internal APIs, databases, and software tools.
• Build and maintain agent evaluation harnesses to benchmark agent accuracy, reliability, and task success across complex, multi-step workflows.
• Analyze and profile agent and inference workloads (e.g., tool-calling latency, multi-agent orchestration) to identify and resolve bottlenecks in compute and latency, using agent observability and profiling tools (e.g., PyTorch Profiler, LangSmith, OpenTelemetry).
• Implement interoperable multi-agent systems using the Agent-to-Agent (A2A) protocol and the Model Context Protocol (MCP), including MCP tools, plugins, and skills.
• Create/maintain CI/CD pipelines for deploying AI Agents to Kubernetes and containerized environments in the cloud.
• Design and implement security controls for autonomous AI Agents, including sandboxing, access control, and mitigation of prompt-injection and tool-misuse risks.
• Mentor engineers on agentic system design paradigms and optimization techniques.

Requirements

WHO YOU ARE

• Technical Degree required (Bachelor’s or Master’s in Computer Science, Statistics, or a related field). Ph.D. is a plus.
• Experience optimizing inference engines (vLLM, TensorRT-LLM) for autonomous AI Agent and LLM workloads.
• Experience developing GenAI applications and AI Agents using one or more frameworks. Microsoft Agent Framework or Google Agent Development Kit preferred.
• Proficiency with Large Language Models (LLMs), including prompt engineering, function calling/tool use, and Chain-of-Thought (CoT) reasoning.
• Experience designing and implementing Retrieval-Augmented Generation (RAG) pipelines, including document ingestion, chunking, embeddings, and vector search.
• Experience building and operating end-to-end systems that automate the testing, evaluation, and deployment of autonomous AI Agents.
• Familiarity with MCP (Model Context Protocol) tools, plugins, and skills, and the Agent-to-Agent (A2A) protocol for building interoperable multi-agent systems.
• Understanding of AI Agent security practices, including sandboxing, access control, and mitigating prompt-injection and tool-misuse risks.
• Familiarity with machine learning frameworks (PyTorch, TensorFlow, scikit-learn, etc).
• Strong scripting and programming skills, with software development skills and a desire to work in a cloud environment.
• 8+ years of relevant industry experience, spanning large-scale data/ML pipelines and/or production Agentic AI systems.
• Outstanding analytical thinking, interpersonal, oral, and written communication skills.
• Ability to prioritize and meet critical project timelines in a fast-paced environment.
• Experience with Kubernetes for orchestrating containerized applications, including deployment and scaling of AI Agent workloads.

NICE TO HAVE

• Experience designing and coordinating Multi-Agent Systems, including collaboration between specialized agents.
• Experience designing and operating agent swarms for distributed, parallelized task execution.
• Experience with agent profiling and evaluation tooling to identify performance regressions and bottlenecks across agent pipelines.
• Hands-on experience with LangChain, LangGraph, or CrewAI.
• Proficiency in TypeScript and/or Rust.
• Demonstrated ability to study and transform data science prototypes into production solutions.
• Knowledge of computer vision and/or signal processing, including techniques for classification and feature extraction.