職缺描述
We‘re looking for an AI Engineer to build next-generation autonomous AI agents leveraging Large Language Models (LLMs). You‘ll design complex agent architectures and behaviors using LangChain, operating across diverse deployment environments from cloud to edge. What You’ll Do 1.Design and develop modular LLM-powered agents with advanced reasoning, planning, and tool orchestration capabilities. 2.Deploy and manage LLMs across diverse environments — from cloud-hosted solutions (OpenAI, Vertex AI, Anthropic) to on-premise/edge setups (llama.cpp, vLLM, Ollama, Nvidia NIM). 3.Track, assess, and troubleshoot agent performance leveraging LangSmith, OpenInference, Datadog LLM Observability, and other monitoring solutions. 4.Develop advanced document processing pipelines that intelligently extract, analyze, and synthesize complex multi-format content. 5.Create advanced retrieval-augmented generation (RAG) systems utilizing vector databases such as FAISS, Milvus, or Pinecone 6.Partner with Solution Architects, ML engineers, DevOps teams, frontend engineers and product stakeholders to transition POCs into robust, production-grade features. 我們正在尋找一位生成式AI夥伴，負責大型語言模型（LLMs）的自動化Agent系統。需要運用 LangChain 架構並能獨立部署，於從雲端到邊緣裝置等多元環境中。 【工作內容】 1.設計並開發模組化的 LLM agents架構，具備可優化演算法的推理、規劃與工具協作能力。 2.部署與管理 LLM 至多種環境 -包含雲端平台（如 OpenAI、Vertex AI、Anthropic）及本地／邊緣端架構（如 llama.cpp、vLLM、Ollama、Nvidia NIM）。 3.熟悉運用 LangSmith、OpenInference、Datadog LLM Observability 等工具，並能提供解決問題方案。 4.開發文件處理流程系統，可解析、提取並彙整多種格式的複雜資訊內容進行提取、分析與整合。 5.開發結合 FAISS、Milvus 或 Pinecone 等向量資料庫的高階檢索增強生成系統(RAG系統)。 6.與解決方案架構師、機器學習工程師、DevOps 團隊、前端工程師及產品相關人員合作，將產品概念驗證（POC）共同推動 成果落地為可用的生產級應用。
收合內容