https://yennj12.js.org/yennj12_blog_V4/tags/cheatsheet/Recent content in Cheatsheet on YennJ12 Engineering BlogHugo -- gohugo.ioen-usTue, 26 May 2026 17:09:24 +0900https://yennj12.js.org/yennj12_blog_V4/posts/ai-fde-essential-guide-part1-zh/Tue, 26 May 2026 16:53:54 +0900https://yennj12.js.org/yennj12_blog_V4/posts/ai-fde-essential-guide-part1-zh/前言 AI Forward Deployed Engineer (FDE) 是連接前沿 AI 技術與生產環境的關鍵角色。不同於傳統的顧問職位，FDE 需要深入客戶環境，從快速原型開發到生產級系統部署，實現可量化的商業價值。本系列文章將深入解析成為優秀 AI FDE 所需的核心技能。 1. Python 生態系統精通 核心語言特性 必須掌握的概念： 1# 生成器與迭代器 - 記憶體效率處理大型數據集 2def data_generator(file_path): 3 with open(file_path, 'r') as f: 4 for line in f: 5 yield process_line(line) 6 7# 異步程式設計 - 提升 I/O 密集型操作效率 8import asyncio 9import aiohttp 10 11async def fetch_embeddings(texts): 12 async with aiohttp.ClientSession() as session: 13 tasks = [get_embedding(session, text) for text in texts] 14 return await asyncio.https://yennj12.js.org/yennj12_blog_V4/posts/ai-fde-essential-guide-part2-zh/Tue, 26 May 2026 16:55:10 +0900https://yennj12.js.org/yennj12_blog_V4/posts/ai-fde-essential-guide-part2-zh/前言 多智慧體系統是現代 AI 應用的重要發展方向，能夠處理複雜的企業級任務。作為 AI FDE，掌握多智慧體框架的設計與實作是核心技能之一。本文將深入探討 LangGraph、CrewAI 等主流框架，以及 Model Context Protocol (MCP) 的實際應用。 1. 多智慧體系統核心概念 基礎架構原理 Agent 核心組件： 感知器 (Perception)：接收與理解環境信息 決策器 (Decision Making)：基於目標與狀態規劃行動 執行器 (Action)：與環境互動執行任務 記憶體 (Memory)：儲存狀態、經驗與知識 協作模式： 1from enum import Enum 2from dataclasses import dataclass 3from typing import List, Dict, Any 4 5class CoordinationPattern(Enum): 6 SEQUENTIAL = "sequential" # 順序執行 7 PARALLEL = "parallel" # 並行執行 8 HIERARCHICAL = "hierarchical" # 階層式管理 9 COLLABORATIVE = "collaborative" # 協作式決策 10 11@dataclass 12class AgentTask: 13 task_id: str 14 description: str 15 agent_id: str 16 dependencies: List[str] 17 priority: int 18 metadata: Dict[str, Any] 19 20class MultiAgentOrchestrator: 21 def __init__(self, coordination_pattern: CoordinationPattern): 22 self.https://yennj12.js.org/yennj12_blog_V4/posts/ai-fde-essential-guide-part3-zh/Tue, 26 May 2026 17:01:52 +0900https://yennj12.js.org/yennj12_blog_V4/posts/ai-fde-essential-guide-part3-zh/前言 企業級 AI 整合與部署是 AI FDE 最具挑戰性的工作之一。需要處理複雜的企業架構、安全合規要求、數據整合與系統可靠性問題。本文將深入探討雲端平台部署策略、企業安全框架、RAG 架構設計與數據管道建構等核心技術。 1. 雲端平台部署策略 Google Cloud Platform (GCP) 深度整合 Vertex AI 生產部署： 1from google.cloud import aiplatform 2from google.cloud.aiplatform import gapic 3import yaml 4 5class GCPAIDeploymentManager: 6 def __init__(self, project_id: str, region: str = "us-central1"): 7 self.project_id = project_id 8 self.region = region 9 10 # 初始化 Vertex AI 11 aiplatform.init( 12 project=project_id, 13 location=region, 14 staging_bucket=f"gs://{project_id}-ml-staging" 15 ) 16 17 def deploy_custom_model(self, model_config: dict): 18 """部署客製化模型到 Vertex AI""" 19 20 # 創建容器映像 21 container_spec = { 22 "image_uri": model_config["container_image"], 23 "env": [ 24 {"name": "MODEL_NAME", "value": model_config["model_name"]}, 25 {"name": "MODEL_VERSION", "value": model_config["version"]}, 26 {"name": "BATCH_SIZE", "value": str(model_config.https://yennj12.js.org/yennj12_blog_V4/posts/ai-fde-essential-guide-part4-zh/Tue, 26 May 2026 17:05:09 +0900https://yennj12.js.org/yennj12_blog_V4/posts/ai-fde-essential-guide-part4-zh/前言 生產環境 AI 系統的監控與最佳化是確保企業 AI 應用成功的關鍵。從模型效能追蹤、基礎設施監控到成本控制，AI FDE 需要建立全方位的可觀測性體系。本文將深入探討 LLM-native 指標設計、分散式監控架構、智能故障診斷與企業級成本最佳化策略。 1. LLM-native 指標與評估體系 核心效能指標設計 LLM 特定指標框架： 1from dataclasses import dataclass 2from typing import Dict, List, Optional, Union 3import numpy as np 4from collections import deque 5import time 6import asyncio 7from enum import Enum 8 9class MetricType(Enum): 10 LATENCY = "latency" 11 THROUGHPUT = "throughput" 12 QUALITY = "quality" 13 COST = "cost" 14 RELIABILITY = "reliability" 15 16@dataclass 17class LLMMetrics: 18 timestamp: float 19 request_id: str 20 model_name: str 21 22 # 效能指標 23 time_to_first_token: float # TTFT - 首個 token 延遲 24 time_per_output_token: float # TPOT - 每個輸出 token 時間 25 total_latency: float 26 tokens_per_second: float 27 28 # 品質指標 29 perplexity: Optional[float] = None 30 bleu_score: Optional[float] = None 31 rouge_score: Optional[Dict[str, float]] = None 32 human_feedback_score: Optional[float] = None 33 34 # 成本指標 35 input_tokens: int = 0 36 output_tokens: int = 0 37 compute_cost: float = 0.https://yennj12.js.org/yennj12_blog_V4/posts/ai-fde-essential-guide-part5-zh/Tue, 26 May 2026 17:09:24 +0900https://yennj12.js.org/yennj12_blog_V4/posts/ai-fde-essential-guide-part5-zh/前言 AI Forward Deployed Engineer 的成功不僅取決於技術能力，更在於與客戶的有效協作與問題解決能力。本系列最終篇將深入探討客戶需求分析、技術溝通策略、專案交付管理，以及從原型到生產的完整實務流程，幫助 AI FDE 實現可量化的商業價值。 1. 客戶需求分析與發現 業務需求挖掘框架 結構化需求分析方法： 1from dataclasses import dataclass 2from typing import Dict, List, Optional, Tuple 3from enum import Enum 4import json 5 6class RequirementPriority(Enum): 7 CRITICAL = "critical" 8 HIGH = "high" 9 MEDIUM = "medium" 10 LOW = "low" 11 12class RequirementType(Enum): 13 FUNCTIONAL = "functional" 14 PERFORMANCE = "performance" 15 SECURITY = "security" 16 COMPLIANCE = "compliance" 17 INTEGRATION = "integration" 18 USABILITY = "usability" 19 20@dataclass 21class BusinessRequirement: 22 requirement_id: str 23 title: str 24 description: str 25 business_value: str 26 success_criteria: List[str] 27 priority: RequirementPriority 28 requirement_type: RequirementType 29 stakeholders: List[str] 30 estimated_impact: str 31 dependencies: List[str] 32 acceptance_criteria: List[str] 33 34class RequirementAnalysisFramework: 35 def __init__(self): 36 self.