AWS SUMMIT

Taipei

Session catalog

Want to learn more about the event?

Explore the session catalog to discover a wide variety of session types, activities, and more. 

65 sessions

10:00 AM CST - Wednesday, Jul 15

  • Keynote
    100 – Foundational
    Session Catalog
    10:00 a.m. Wednesday, Jul 15
    Wednesday, Jul 15
    AI 代理（AI Agent）的時代已經來臨。在這場 Keynote 中，AWS 將揭示如何讓代理式 AI（Agentic AI）真正落地——從建構、部署到大規模擴展，AWS 正在為開發者移除每一道障礙。你將聽到來自業界先驅的真實故事，了解領先企業如何運用 AI 與代理式架構實現突破性成果；同時見證全新開發者工具與流程如何讓每位開發者以 AI 為核心加速創新。無論你是正在打造下一個突破性應用的開發者，還是引領組織數位轉型的 IT 決策者，這場 Keynote 將為你揭開 AI 驅動未來的完整路徑。(The era of AI Agents has arrived. In this Keynote, AWS will reveal how to make Agentic AI truly production-ready — from building and deploying to scaling at enterprise level, AWS is removing every barrier for developers. You will hear real stories from industry pioneers and learn how leading enterprises are leveraging AI and agentic architectures to achieve breakthrough results; while witnessing how new developer tools and workflows empower every developer to accelerate innovation with AI at the core. Whether you are a developer building the next breakthrough application, or an IT decision-maker leading your organization's digital transformation, this Keynote will unveil the complete path to an AI-driven future.)
    • Wednesday, Jul 1510:00 AM - 11:00 AM CST
      3F 大會堂

    , Managing Director, Taiwan, Amazon Web Services, Inc.

    , Director, Next Generation Developer Experience, GenAI GTM, Amazon Web Services, Inc.

    , Enterprise Product Lead for Codex, Open AI

    , 代表取締役社長CEO, アスクル株式会社

    Session Type: Keynote

11:00 AM CST - Wednesday, Jul 15

  • Automation
    Agentic AI
    DevOps
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    Student
    System Administrator
    Tech Explorer
    Breakout session
    Session Catalog
    200 – Intermediate
    AI/ML
    11:00 a.m. Wednesday, Jul 15
    Wednesday, Jul 15
    探索 AI-DLC 如何將 AI 作為軟體開發全生命週期中的核心協作者，透過AI驅動流程、持續上下文與人機協作模式，串聯 Inception、Construction 與 Operations 三大階段，加速開發效率、提升品質，並打造更佳的開發者體驗。(Discover how AI-DLC transforms software engineering by positioning AI as a core collaborator across the entire development lifecycle. Learn how AI-driven workflows, persistent context, and human oversight work together through the Inception, Construction, and Operations phases to accelerate delivery, improve quality, and enhance developer experience. )
    • Wednesday, Jul 1511:30 AM - 12:15 PM CST
      Room 201DEF

    , Cloud Architect, AWS

    , Assoc. Delivery Consultant, AWS

    Session Topic: AI/ML
    Session Type: Breakout session
    Level: 200 – Intermediate
  • Business Intelligence
    Automation
    Learning from Amazon
    Academic / Researcher
    Advisor / Consultant
    Business Executive
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Sales / Marketing
    Solution / Systems Architect
    Tech Explorer
    Amazon Quick Suite
    Amazon Quick
    Breakout session
    Session Catalog
    200 – Intermediate
    AI/ML
    Analytics
    Architecture
    11:00 a.m. Wednesday, Jul 15
    Wednesday, Jul 15
    規模化推行企業 AI，遠比想像中複雜。亞馬遜內部跨多個事業部、將 Quick 導入數千名員工的歷程中，曾面對數據整合的困境、架構設計的取捨，以及跨部門協作的摩擦。我們將直接分享實戰心得，協助你建立一套真正可落地的企業 AI 推行框架。(Go behind the scenes of Amazon's internal Quick Suite deployment across thousands of employees and multiple organizations. Amazon teams share real challenges in data integration, orchestration, and change management, along with what they got wrong. Leave with practical frameworks for enterprise AI adoption that any organization can apply.)
    • Wednesday, Jul 1511:30 AM - 12:15 PM CST
      3F 大會堂

    , Sr Mgr Business Insights, Amazon

    Session Topic: AI/ML, Analytics, Architecture
    Session Type: Breakout session
    Level: 200 – Intermediate
  • Business Intelligence
    Business Executive
    Developer / Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    Tech Explorer
    AWS Billing and Cost Management
    Breakout session
    Session Catalog
    200 – Intermediate
    Architecture
    Business Applications
    11:00 a.m. Wednesday, Jul 15
    Wednesday, Jul 15
    以 Cost Intelligence Dashboard 在 AWS 上建構統一的多雲成本儀表板，整合來自 AWS、Azure 與 GCP 的 FOCUS 標準成本資料，實現跨雲支出的全面可視化與分析。(Build a unified multi-cloud cost dashboard on AWS using Cost Intelligence Dashboard, integrating FOCUS-standard cost data from AWS, Azure, and GCP for cross-cloud spend visibility and analysis.)
    • Wednesday, Jul 1511:30 AM - 12:15 PM CST
      Room 101BC

    , Technical Account Manager, AWS

    , Technical Account Manager, AWS

    Session Topic: Architecture, Business Applications
    Session Type: Breakout session
    Level: 200 – Intermediate
  • Automation
    Agentic AI
    DevOps
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    AWS DevOps Agent
    Breakout session
    Customer story
    Session Catalog
    300 – Advanced
    AI/ML
    Architecture
    DevOps & Developer Productivity
    11:00 a.m. Wednesday, Jul 15
    Wednesday, Jul 15
    瞭解 AWS DevOps Agent 如何整合遙測資料、程式碼與部署紀錄，自主追查問題根因、產出修復方案，並主動預防未來事故的發生。(Learn how AWS DevOps Agent correlates telemetry, code, and deployment data to autonomously investigate root causes, generate mitigation plans, and proactively prevent future incidents.)
    • Wednesday, Jul 1511:30 AM - 12:15 PM CST
      Room 201ABC

    , Solutions Architect, AWS

    , Staff SRE, Trendmicro

    Session Topic: AI/ML, Architecture, DevOps & Developer Productivity
    Session Type: Breakout session
    Level: 300 – Advanced
    Session Features: Customer story
  • Amazon Elastic Kubernetes Service (Amazon EKS)
    Workshop
    Hands-on
    Laptop required
    Session Catalog
    Apply to attend
    11:00 a.m. Wednesday, Jul 15
    Wednesday, Jul 15
    在這個工作坊中，您將學習如何在 Amazon EKS 上建構、部署並運維正式環境的 AI Agent。 您將深入探索兩種不同的架構方式——使用 vLLM、Langfuse、Milvus 與 MCP 的自建基礎設施，以及整合 Amazon Bedrock 與 Amazon Agentcore 的 AWS 託管服務方案。透過實作練習，您將部署單一 Agent 與多 Agent 系統（使用 Strands SDK）、實作端到端的可觀測性、管理 Agent 記憶體、設定安全的工具存取權限，並透過 A2A 協定實現 Agent 之間的無縫通訊。無論您想完整掌控自己的 AI 技術堆疊，還是善用全託管的 AWS 服務，完成本工作坊後，您都將具備將 Agentic AI 工作負載帶入正式環境的實戰能力。
    • Wednesday, Jul 1511:30 AM - 1:30 PM CST
      Room 202

    , Sr. Specialist Solutions Architect, AI & Engineering, AWS

    Session Type: Workshop
    Session Features: Hands-on, Laptop required

1:00 PM CST - Wednesday, Jul 15

  • Amazon Quick
    Breakout session
    Session Catalog
    1:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    AI Agent 不再只是概念——本場次以 Amazon Quick 為實例，帶你快速掌握 MCP（Model Context Protocol）與 Skills 的核心架構，了解如何串接多重工作流並精準定位 Agent 的角色。從零開始上手基本概念，再到中等複雜度的流程實作，讓你親身體驗 AI 如何為日常工作加速。(AI Agents are no longer just a concept — this session uses Amazon Quick as a live example to help you quickly grasp the core architecture of MCP (Model Context Protocol) and Skills, and understand how to orchestrate multiple workflows while precisely defining the Agent's role. Starting from foundational concepts from scratch, then progressing to moderate-complexity workflow implementation, you will experience firsthand how AI accelerates everyday work.)
    • Wednesday, Jul 151:30 PM - 2:15 PM CST
      Room 102

    , Technical Trainer, AWS

    Session Type: Breakout session
  • Agentic AI
    Manufacturing & Industrial
    Business Executive
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Sales / Marketing
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Amazon Bedrock
    Lightning talk
    Community-led
    Session Catalog
    200 – Intermediate
    AI/ML
    Business Applications
    1:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    在許多製造業公司，ERP 料號查詢和產線回報至今仍仰賴採購、業務支援和工程師每天手動翻 Excel、開 ERP 畫面、跨部門問來問去。更大的隱性成本是：每次有人離職，新人就得重新學習複雜的 ERP 操作，或接手一堆只有前任看得懂的 Excel — 營運知識最終被鎖死在工具的學習曲線裡。傳統的解法 — 導入 BI 工具或委外開發內部系統 — 通常要兩個月加上一整個專案團隊。 我們團隊在客戶工廠裡走了一條不同的路：使用代理式編程工具（Agentic Coding Tools）搭配 Amazon Bedrock 上的 Claude 模型，從四個製造廠區的 ERP 匯出資料開始，第一個可用版本在一到兩小時內就能完成 — 而且流程簡單到連資淺成員也能端到端跑完。這不是 Demo：系統已經在製造業客戶的正式環境上線，每天由採購、業務支援和現場人員實際使用。 這場 Lightning Talk 會帶你走過四件事：(1) 當非技術人員可以用自然語言直接查詢料號資料時，互動方式有什麼改變；(2) 為什麼我們選擇 Amazon Bedrock 而非公開的 LLM API — 每一次 LLM 呼叫都留在客戶自己的 AWS 帳號內，模型不會拿客戶資料做訓練、資料不會跨出客戶邊界，這是製造業客戶在合規與資安審查下的決定性因素；(3) 我們團隊長期發展出的代理式編程互動風格 — 能觸發長時間自主執行、一次到位，不需要反覆來回修正；(4) 最後一哩路 — 讓公司裡每一位員工都能自己登入使用。
    • Wednesday, Jul 151:30 PM - 1:55 PM CST
      2F Developer Community Zone

    , Managing Director, Kyklosify

    Session Topic: AI/ML, Business Applications
    Session Type: Lightning talk
    Level: 200 – Intermediate
    Session Features: Community-led
  • Automation
    Agentic AI
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Amazon Bedrock
    Kiro
    Breakout session
    Session Catalog
    200 – Intermediate
    AI/ML
    Architecture
    1:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    AWS AI Engineering 以 Prompt、Context、Harness 三層心智模型，重新定義 AI Agent 的開發模式：Prompt 打造清晰指令、Context 透過 MCP 與 Skills 動態組裝對的知識、Harness 以編排、工具、記憶、可觀測性、評測、恢復六大支柱建構完整控制系統。應用場景橫跨舊系統現代化、Agentic 工作流自動化與企業級 AI 部署。搭配 Amazon Kiro 加速 Spec-Driven 開發、Amazon Bedrock AgentCore 作為 Agent 運行底座，讓團隊跨越 PoC 鴻溝，真正把 AI Agent 推上 Production。(AWS AI Engineering redefines the AI Agent development paradigm with a three-layer mental model of Prompt, Context, and Harness: Prompt crafts clear instructions; Context dynamically assembles the right knowledge through MCP and Skills; and Harness builds a complete control system with six pillars — orchestration, tools, memory, observability, evaluation, and recovery. Application scenarios span legacy system modernization, Agentic Workflow automation, and enterprise-grade AI deployment. Paired with Amazon Kiro to accelerate Spec-Driven Development and Amazon Bedrock AgentCore as the Agent runtime foundation, teams can bridge the PoC gap and truly push AI Agents into Production.)
    • Wednesday, Jul 151:30 PM - 2:15 PM CST
      Room 201ABC

    , Solutions Architect, AWS

    , Prototyping SA, AWS

    Session Topic: AI/ML, Architecture
    Session Type: Breakout session
    Level: 200 – Intermediate
  • Global Infrastructure
    Developer / Engineer
    DevOps Engineer
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Breakout session
    Session Catalog
    300 – Advanced
    Architecture
    Cloud Operations
    Hybrid Cloud
    Networking
    1:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    AWS 台灣區域（Region）自啟用以來即配備三座可用區域（Availability Zones），實現基礎設施韌性的最高標準。本場次以此為起點，深入探討如何在 AWS 上設計高韌性架構，從單區域多 AZ 部署到跨區域容錯，提供具體的架構決策框架與設計模式。 接著，我們邀請 TheKeeper 分享其 AI 兒童陪伴產品「巧虎伴伴」的完整技術旅程：如何透過 Amazon Bedrock 串接多種大型語言模型（包含 Claude 與 OpenAI），打造能理解兒童情緒、即時回應的 AI 角色互動體驗；同時運用 AWS 韌性架構，確保百萬級兒童用戶在尖峰時段仍享有低延遲、不中斷的陪伴服務。從雲端遷移決策、Bedrock 模型編排、IP 內容保護，到面向亞太市場的全球擴展架構規劃——一次完整呈現。(Since its launch, the AWS Taiwan Region has been equipped with three Availability Zones, achieving the highest standard of infrastructure resilience. This session takes that as a starting point to dive deep into designing highly resilient architectures on AWS — from single-region multi-AZ deployments to cross-region fault tolerance — providing concrete architectural decision frameworks and design patterns. Next, we invite The Keeper to share the complete technical journey of their AI children's companion product "Qiaohu Companion" (巧虎伴伴): how they leverage Amazon Bedrock to orchestrate multiple large language models (including Claude and OpenAI) to create an AI character interaction experience that understands children's emotions and responds in real time; while utilizing AWS resilient architecture to ensure millions of child users still enjoy low-latency, uninterrupted companion services during peak hours. From cloud migration decisions, Bedrock model orchestration, and IP content protection, to global expansion architecture planning targeting the Asia-Pacific market — presented as a complete end-to-end story.)
    • Wednesday, Jul 151:30 PM - 2:15 PM CST
      Room 101BC

    , Sr. Solutions Architect, AWS

    , Tech Lead, Voice Keeper Co., Ltd.

    Session Topic: Architecture, Cloud Operations, Hybrid Cloud, Networking
    Session Type: Breakout session
    Level: 300 – Advanced
  • Serverless
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Breakout session
    Session Catalog
    300 – Advanced
    Architecture
    Serverless Compute & Containers
    1:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    Migrating a live banking platform to microservices is never just a technical challenge. Grounded in a real modernisation project rebuilding a bank's omnichannel platform on AWS for 20+ million customers, this session walks through four challenges and the hard lessons from each: nailing down requirements when different teams disagree and gates surface late (user stories before LLDs, contract-first and mock-first delivery); deciding where to draw service boundaries (Bounded Context), and earning the right to split; making new and old systems coexist seamlessly (the Strangler Fig pattern); and increasing delivery velocity in the agentic AI age (CI/CD from day one, engineering standards, tackling the code review bottleneck, and steering docs as the new code). Leave with patterns you can apply immediately, whether you are planning your migration or already mid-journey. 將運行中的銀行平台遷移至微服務，挑戰從不只是技術層面。本場次以一個真實的現代化專案為基礎，該專案在 AWS 上為超過 2,000 萬名客戶重建銀行的全通路平台。session 將依序剖析四大挑戰及各自的實戰教訓：在各團隊需求分歧、關卡又遲遲浮現時如何釘牢需求（先寫 User Story 而非 LLD、合約優先與 Mock 優先的交付方式）；如何劃定服務邊界（Bounded Context)，以及「贏得拆分的資格」；如何讓新舊系統無縫共存（Strangler Fig 模式）；以及在代理式AI（Agentic AI）時代如何提升交付速度（從第一天就建立 CI/CD、工程標準、破解程式碼審查瓶頸，以及作為新程式碼的 Steering Docs）。無論您正在規劃遷移或已身處旅程之中，都能帶走可立即落地的實用模式。
    • Wednesday, Jul 151:30 PM - 2:15 PM CST
      Room 201DEF

    , Sr. Delivery Consultant-AppMod, AWS

    Session Topic: Architecture, Serverless Compute & Containers
    Session Type: Breakout session
    Level: 300 – Advanced
  • Agentic AI
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Amazon Bedrock
    Breakout session
    Customer story
    Session Catalog
    200 – Intermediate
    AI/ML
    Architecture
    1:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    生成式 AI 改變了您的架構方式。智慧代理正在改變您的業務運作。本次演講帶您回顧 AWS 最具影響力的 Agentic AI、生成式 AI 與機器學習重磅發布，讓您精準掌握產業轉折，明確組織的下一步行動。(Generative AI changed how you build. Agentic AI changes how your business runs. In this session, we recap the most significant agentic AI, generative AI, and ML announcements from AWS so you know exactly what shifted and exactly what to do next.)
    • Wednesday, Jul 151:30 PM - 2:15 PM CST
      3F 大會堂

    , Solutions Architect, AWS

    , Director of APAC Partner Alliance, TrendAI

    Session Topic: AI/ML, Architecture
    Session Type: Breakout session
    Level: 200 – Intermediate
    Session Features: Customer story
  • Network & Infrastructure Security
    Agentic AI
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    AWS Security Agent
    Lightning talk
    Community-led
    Session Catalog
    200 – Intermediate
    AI/ML
    Security, Compliance, & Identity
    1:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    AWS Security Agent是一個自動化、資安面向的AI Agent。強大的自動滲透測試功能可以讓你輕鬆找出網站的安全漏洞。但 AI Agent的不確定性，也帶來一些風險。儘管 AWS Security Agent已經有很多預設安全防護，避免滲透測試對目標網站造成傷害。但在設定測試時，也有一些步驟可以注意，讓我們的 Security Agent 滲透測試更完美，更安全。 在這個環節會介紹幾個用戶在設定 Security Agent 時經常忽略的安全隱患，以及對應的防護策略︰ 1. 低估了滲透測試的破壞性，以 production 環境作為測試目標。 2. 過份依賴 Security Agent 的安全護欄，沒有仔細定義甚麼路徑是不可以攻擊。令滲透測試存在對第三方網站攻擊的風險。 3. 低估測試報告的詳細程度，沒有正確設定 agent space 的存取權限，以致敏感資料外洩。
    • Wednesday, Jul 151:55 PM - 2:20 PM CST
      2F Developer Community Zone

    , AWS Security Hero, AWS Security Hero

    Session Topic: AI/ML, Security, Compliance, & Identity
    Session Type: Lightning talk
    Level: 200 – Intermediate
    Session Features: Community-led

2:00 PM CST - Wednesday, Jul 15

  • Workshop
    Hands-on
    Laptop required
    Session Catalog
    Apply to attend
    2:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    AI Agent 正在改變軟體開發的面貌，而最好的學習方式就是動手做！ Agentic Football Cup 已在全球完成 100+ 場活動（涵蓋印度、澳洲、拉丁美洲、歐洲及美國），開發者評分 4.8/5，現在輪到台北了！ 你負責撰寫策略，你的 AI Agent 上場比賽 — 每隊 5 位自主 AI 代理人，每 2 秒做出傳球、射門、壓迫、防守等決策，一旦哨聲響起，完全無需人為介入。 這不是觀看的 Demo，而是你親手建造並部署的工作系統。 你將學到：Amazon Bedrock、AgentCore 與 Strands Agents SDK — 與企業級 Agentic 系統相同的技術堆疊。你將實際理解 Multi-Agent Orchestration、即時狀態管理、Tool Use 與 Guardrails。 【誰適合參加】各程度的開發人員 — 無需 AI/ML 經驗。 適合：開發人員、解決方案架構師、技術主管、資料科學家。 需具備基礎程式能力（Python 為佳）。
    • Wednesday, Jul 152:00 PM - 5:00 PM CST
      Room 103

    , Sr. Partner Solution Architect, AWS

    Session Type: Workshop
    Session Features: Hands-on, Laptop required
  • Artificial Intelligence
    Business Applications
    Business Intelligence
    Generative AI
    Agentic AI
    Advisor / Consultant
    Business Executive
    IT Professional / Technical Manager
    Sales / Marketing
    Student
    Tech Explorer
    Amazon Quick
    Workshop
    100 – Foundational
    Hands-on
    Laptop required
    Session Catalog
    Apply to attend
    2:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    在資訊爆炸的時代，每天的工作充斥著大量文件整理、資料分析、報告撰寫與跨部門溝通。Amazon Quick 是 AWS 推出的全方位 AI 智慧工作平台，整合生成式 AI 搜尋、深度研究、自動化工作流程與智能文件產出，讓您無需撰寫任何程式碼，即可大幅提升日常工作效率。本工作坊將帶您實際操作 Amazon Quick 的核心功能：透過 Quick Research 對大量文件進行深度研究與洞察分析；利用 Flows 自動化重複性工作流程（如報告產出、資料彙整、審核通知）；運用 Spaces 建立團隊共享知識庫，讓 AI 助理即時回答業務問題；以及直接在對話中產出 Word、PowerPoint、Excel 等專業文件，省去反覆切換工具的時間。無論您是需要快速產出研究報告的研究人員、希望優化團隊協作的專案經理，還是想提升日常行政效率的營運人員，本工作坊都將為您開啟 AI 驅動的全新工作方式。 13:30–14:00 報到與交流 14:00–14:10 Agentic AI 如何重塑企業工作模式 - 高效管理之道。 14:10–14:30 設定 Amazon Quick 桌面版 14:30–15:00 Lab1: AI 時代的時間管理 - 製作 Email AI 助理，拯救您的時間 15:00-15:10 Break: 休息 15:10-15:30 Lab2: 打造高效工作助理 -  善用 Chat Agent, Knowledge Base 自動排定重要活動 15:30–16:00 Lab3: 打好團體戰 - 善用 Skill, Memory 打造更聰明的你和代理人們。總結 16:00–16:10 Break: 休息 16:10–16:30 進階功能分享(Browser Automation, Deep Analysis...等), QA
    • Wednesday, Jul 152:00 PM - 4:30 PM CST
      Room 202

    , Technical Account Manager, AWS

    Session Topic: Artificial Intelligence, Business Applications
    Session Type: Workshop
    Session Features: Hands-on, Laptop required
  • Serverless
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    Student
    System Administrator
    Tech Explorer
    Amazon Aurora
    Amazon Simple Email Service (Amazon SES)
    Amazon Simple Queue Service (Amazon SQS)
    AWS Lambda
    Lightning talk
    Community-led
    Session Catalog
    200 – Intermediate
    Serverless Compute & Containers
    2:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    過去在 AWS Educate 校園大使計畫的日常營運背後，存在著大量重複性的人工作業與跨團隊協作痛點，為此成員自發組成 Dev Team，結合軟體工程實踐與 AWS Serverless 服務，打造對外的官方網站與對內的自動化信件系統。官方網站以 Next.js 和 Tailwind CSS 開發，作為計畫介紹與成果展示的統一入口；自動化信件系統則以 AWS Lambda 搭配 Amazon SQS 驅動微服務事件處理，Amazon Aurora Serverless v2 儲存信件寄送任務，Amazon SES 實現百至千人規模的客製化大量信件寄送。這場分享將帶你從真實營運痛點出發，深入了解我們如何一步步完成系統設計、開發與上線，並透過雲端與軟體工程技術推動團隊營運自動化。(Behind the daily operations of the AWS Educate Student Ambassador program, there existed a significant amount of repetitive manual work and cross-team collaboration pain points. To address this, members proactively formed a Dev Team, combining software engineering practices with AWS Serverless services to build a public-facing official website and an internal automated email system. The official website was developed with Next.js and Tailwind CSS, serving as a unified portal for program introduction and achievement showcases. The automated email system uses AWS Lambda paired with Amazon SQS to drive microservice event processing, Amazon Aurora Serverless v2 to store email delivery tasks, and Amazon SES to enable customized mass email delivery at a scale of hundreds to thousands of recipients. This session will take you from real operational pain points to an in-depth understanding of how we completed system design, development, and launch step by step, leveraging cloud and software engineering technologies to drive team operational automation.)
    • Wednesday, Jul 152:20 PM - 2:45 PM CST
      2F Developer Community Zone

    , AWS Educate Cloud Ambassador, AWS Educate

    , AWS Educate Cloud Ambassador

    , AWS Educate Cloud Ambassador

    , AWS Educate Cloud Ambassador

    , Cloud Ambassador, AWS Educate TW

    Session Topic: Serverless Compute & Containers
    Session Type: Lightning talk
    Level: 200 – Intermediate
    Session Features: Community-led
  • Academic / Researcher
    Advisor / Consultant
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Sales / Marketing
    Solution / Systems Architect
    Student
    System Administrator
    Tech Explorer
    Venture Capitalist
    Exam prep
    Session Catalog
    100 – Foundational
    AI/ML
    2:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    AIF-C01 為 AWS 入門級人工智慧認證，不要求程式開發能力或深度技術架構知識。該認證旨在驗證考生是否具備清楚闡述人工智慧、機器學習及生成式 AI 於 AWS 平台上之應用概念的能力。本認證適合以下角色：業務開發人員、專案經理、技術顧問，以及希望取得 AI 基礎能力證明之初階工程師。本課程將系統性地帶領學員掌握 AIF-C01 考試範圍，包含各領域配分比重、核心概念重點，以及高頻考點解析，協助學員以最有效率的方式通過認證考試。(AIF-C01 is an AWS entry-level artificial intelligence certification that does not require programming skills or in-depth technical architecture knowledge. This certification is designed to validate whether candidates can clearly articulate the concepts and applications of artificial intelligence, machine learning, and generative AI on the AWS platform. It is suitable for the following roles: business developers, project managers, technical consultants, and junior engineers seeking to obtain proof of foundational AI competency. This course will systematically guide learners through the AIF-C01 exam scope, including domain weight distribution, core concept highlights, and high-frequency exam point analysis, helping learners pass the certification exam in the most efficient way possible.)
    • Wednesday, Jul 152:30 PM - 3:15 PM CST
      Room 102

    , Technical Trainer, AWS

    Session Topic: AI/ML
    Session Type: Exam prep
    Level: 100 – Foundational
  • Agentic AI
    Academic / Researcher
    Advisor / Consultant
    Business Executive
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    Student
    System Administrator
    Tech Explorer
    Kiro
    Breakout session
    Session Catalog
    200 – Intermediate
    AI/ML
    2:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    AI 生成程式碼很快，但如何確保品質、一致性與可追溯性？本場次以獨立遊戲開發為例，展示 Kiro IDE 如何透過 Steering、Spec、Hooks、Powers 四大機制，讓 AI Agent 從「隨機生成」進化為「有章法的工程交付」。適合所有想瞭解 AI-native IDE 如何改變開發流程的人。(AI generates code fast — but how do you ensure quality, consistency, and traceability? Using indie game development as an example, this session demonstrates how Kiro IDE leverages four key mechanisms — Steering, Spec, Hooks, and Powers — to evolve AI Agents from "random generation" to "disciplined engineering delivery." Ideal for anyone who wants to understand how an AI-native IDE can transform the development workflow.)
    • Wednesday, Jul 152:30 PM - 3:00 PM CST
      Room 201ABC

    , Senior Solutions Architect, AWS

    Session Topic: AI/ML
    Session Type: Breakout session
    Level: 200 – Intermediate
  • Advisor / Consultant
    Business Executive
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Sales / Marketing
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Amazon Bedrock
    Breakout session
    AWS Partners
    Session Catalog
    200 – Intermediate
    AI/ML
    2:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    十年前，Cloud 重塑了整個產業的遊戲規則；今天，GenAI 工具正在做同一件事 —— 只是這次速度更快、留給觀望者的時間更少。 本議程將從顧問與客戶面向，呈現企業在 AI 時代的轉型路徑。首先，我們將分析主流 GenAI 工具如何改變客戶對服務的期待，從交付模式到顧問價值定義，說明哪些能力依然是護城河、哪些正在被重寫。其次，分享 Nextlink 如何將 Amazon Bedrock 導入自身顧問工作流程 —— 從需求訪談、架構建議到客戶報告產製， 全面提升內部交付效率，並將這套經過實戰驗證的流程，轉化為客戶可感受到的服務品質提升與響應速度。第三，以 Spec-Driven Development 框架為基礎，分享在多產業執行短週期 AI 落地專案的第一手經驗，分析所面對到的疑難雜症。
    • Wednesday, Jul 152:30 PM - 3:00 PM CST
      3F 大會堂

    , 副理, Nextlink Technology

    Session Topic: AI/ML
    Session Type: Breakout session
    Level: 200 – Intermediate
    Session Features: AWS Partners
  • Agentic AI
    Government
    Manufacturing & Industrial
    Academic / Researcher
    Advisor / Consultant
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    Student
    System Administrator
    Tech Explorer
    Breakout session
    AWS Partners
    Session Catalog
    100 – Foundational
    AI/ML
    Hybrid Cloud
    2:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    企業渴望加速推動 AI 與現代化工作負載，但多數應用程式仍受限於地端。傳統做法得先經過系統重構 (Refactoring) 與平台移轉 (Replatforming)，曠日費時，難以即時取用雲端資源。 Nutanix Cloud Clusters (NC2) on AWS 讓您免重構、直接將地端環境延伸上雲。本講座將說明 NC2 如何簡化工作負載遷移、最佳化成本，並與 AWS 服務整合。您也將瞭解 NC2 在儲存、網路與災難備援上的進階功能，以及如何透過統一管理平台簡化 IT 營運。 立即加入本場 Nutanix 專題講座。我們將一同探索傳統基礎架構上雲的安全途徑，讓您順利接軌現代化 AI 與各項雲端服務，推動業務創新！
    • Wednesday, Jul 152:30 PM - 3:00 PM CST
      Room 201DEF
    Session Topic: AI/ML, Hybrid Cloud
    Session Type: Breakout session
    Level: 100 – Foundational
    Session Features: AWS Partners
  • Academic / Researcher
    Advisor / Consultant
    Business Executive
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Sales / Marketing
    Solution / Systems Architect
    Student
    System Administrator
    Tech Explorer
    Breakout session
    AWS Partners
    Session Catalog
    200 – Intermediate
    Architecture
    2:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    中華電信將分享導入 AWS 雲端韌性的實務經驗，說明如何依 RTO/RPO 與成本選擇 Backup & Restore 到 Active-Active 各級架構，並以監控、自動切換與災難演練，將韌性轉化為可驗證、可治理的營運能力。
    • Wednesday, Jul 152:30 PM - 3:00 PM CST
      Room 101BC

    , Director, Chunghwa Telecom

    Session Topic: Architecture
    Session Type: Breakout session
    Level: 200 – Intermediate
    Session Features: AWS Partners
  • Agentic AI
    Financial Services
    Business Executive
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Sales / Marketing
    Solution / Systems Architect
    Tech Explorer
    Amazon Lightsail
    Lightning talk
    Customer story
    Session Catalog
    200 – Intermediate
    AI/ML
    Business Applications
    2:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    Blave 是一間專為現代交易者打造量化 AI 基礎設施的公司，將研究、回測、監控與交易自動化結合 AI，實現 No-coding 量化交易工作系統。本次分享將透過 BlaveClaw 的實戰經驗，分享如何利用 AWS Lightsail 解決 AI Agent 部署的最後一哩路，讓使用者能夠一鍵建立專屬的交易助理，並透過獨立的雲端運行環境，實現穩定、安全且 24/7 持續運作的自動化交易流程，真正將交易想法化為現實。
    • Wednesday, Jul 152:45 PM - 3:10 PM CST
      2F Developer Community Zone
    Session Topic: AI/ML, Business Applications
    Session Type: Lightning talk
    Level: 200 – Intermediate
    Session Features: Customer story

3:00 PM CST - Wednesday, Jul 15

  • Student
    Lightning talk
    Community-led
    Session Catalog
    100 – Foundational
    3:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    從學生時期加入 AWS Educate 雲端校園大使種子計畫開始，我們透過 AWS Community 接觸雲端技術、參與技術交流、投入開源貢獻，並將 AWS Services 應用於實務場景。本場 Session 將分享這段 Community 經驗如何影響我們從學生到工程師的技術學習、實作累積與職涯發展
    • Wednesday, Jul 153:10 PM - 3:35 PM CST
      2F Developer Community Zone

    , Cloud Engineer, SHOPLINE

    , Solutions Architect, iKala

    , Software Engineer, TSMC

    Session Type: Lightning talk
    Level: 100 – Foundational
    Session Features: Community-led
  • Gamified learning
    Session Catalog
    100 – Foundational
    3:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    想用更有趣的方式學 AWS？本場次完整介紹 AWS Skill Builder 的學習生態系——從官方課程、講師授課到沙箱實驗環境，再到卡牌收集、密室逃脫與 RPG 闖關等遊戲式學習模式，以及企業模擬情境的互動實作。無論你是剛起步或想探索多元學習路徑，這場次帶你解鎖全方位的 AWS 學習體驗。
    • Wednesday, Jul 153:15 PM - 4:00 PM CST
      Room 102

    , Technical Trainer, AWS

    , Technical Trainer, AWS

    Session Type: Gamified learning
    Level: 100 – Foundational
  • Automation
    Agentic AI
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Kiro
    Breakout session
    Customer story
    Session Catalog
    300 – Advanced
    AI/ML
    3:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    Kiro 的 AI-DLC 以智慧化自動工作流程，重新定義現代開發模式，讓快速迭代與跨職能協作成為常態。這套新世代開發生命週期整合了各司其職的 AI 代理人：創意代理人策劃行銷活動、資料代理人優化即時分析洞察、支援代理人深化客戶互動體驗。應用場景橫跨敏捷產品創新、營運效率提升與高度個人化的使用者體驗。從策略規劃到執行落地，代理人全面分工，讓團隊以前所未有的靈活度加速上市、持續催生突破性創新。Kiro 協助組織以 AI 驅動的彈性解決方案，將挑戰轉化為機會。(Kiro's AI-DLC redefines development through automated, intelligent workflows that enable rapid iteration and cross-functional collaboration. This next-gen lifecycle integrates AI agents built for distinct roles: a Creative Agent crafts marketing campaigns, a Data Agent optimizes real-time analytics, and a Support Agent enhances customer engagement. Use cases span agile product innovation, operational efficiency, and hyper-personalized user experiences. With agents covering everything from strategic planning to hands-on execution, teams gain the agility to accelerate time-to-market and drive breakthrough innovation. Kiro helps organizations turn challenges into opportunities through adaptive, AI-driven solutions.)
    • Wednesday, Jul 153:15 PM - 4:00 PM CST
      Room 201ABC

    , Senior Solutions Architect, AWS

    , Sr. AI Engineer, 新加坡商甯寶數位科技有限公司

    Session Topic: AI/ML
    Session Type: Breakout session
    Level: 300 – Advanced
    Session Features: Customer story
  • Global Infrastructure
    Network & Infrastructure Security
    Telecommunications
    Developer / Engineer
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    AWS Cloud WAN
    Breakout session
    Customer story
    Session Catalog
    400 – Expert
    Architecture
    Cloud Operations
    Hybrid Cloud
    Networking
    3:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    本議程說明如何以 AWS Cloud WAN 建立高可控、高靈活性的多雲動態路由架構，並邀請 Nokia 分享電信核心網路的實務案例：說明如何以代理式 AI 接入控制，依網路容量狀況動態調節私有雲與 AWS 公有雲核心網路間的流量切換節奏，並示範容量彈性擴展、AI 過載保護與服務健康監控，達成電信級的高韌性災難復原。(This session explains how AWS Cloud WAN delivers highly controllable, highly flexible multicloud dynamic routing, and features a Nokia case study on telecom core network resilience: how agentic AI-based admission control dynamically paces traffic switchover between private-cloud and AWS public-cloud cores based on live capacity, with capacity scale-out, AI-driven overload protection, and service health monitoring to achieve carrier-grade resilient disaster recovery.)
    • Wednesday, Jul 153:15 PM - 4:00 PM CST
      Room 101BC

    , Cloud Support Engineer, AWS

    , Presales, Nokia

    Session Topic: Architecture, Cloud Operations, Hybrid Cloud, Networking
    Session Type: Breakout session
    Level: 400 – Expert
    Session Features: Customer story
  • Agentic AI
    Kubernetes
    Financial Services
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Amazon Elastic Kubernetes Service (Amazon EKS)
    Breakout session
    Customer story
    Session Catalog
    300 – Advanced
    AI/ML
    Architecture
    Cloud Operations
    Kubernetes
    3:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    當 AI 輔助開發在 BitoGroup 工程團隊擴散,平台團隊面對熟悉的岔路:讓每個團隊各自架 OpenClaw,還是打造治理平台？BitoGroup 選擇後者,在 Amazon EKS 上 ship 出多租戶 OpenClaw 平台,具備團隊層級隔離、預算可視化、Scale-to-Zero 成本模型,讓閒置團隊不產生花費。了解背後的決策框架、AWS-native 與 cloud-agnostic 路線的取捨,以及適用於任何多租戶開發者平台的設計模式,並參考 BitoGroup 協力發布至 aws-samples 的實作。(As AI-assisted development spreads across BitoGroup's engineering teams, the platform team faces a familiar crossroads: let each team spin up their own OpenClaw instance, or build a governance platform? BitoGroup chose the latter, shipping a multi-tenant OpenClaw platform on Amazon EKS with team-level isolation, budget visibility, and a Scale-to-Zero cost model that ensures idle teams incur no costs. Learn about the decision framework behind this choice, the trade-offs between AWS-native and cloud-agnostic approaches, and design patterns applicable to any multi-tenant developer platform — with reference to the implementation BitoGroup co-published to aws-samples.)
    • Wednesday, Jul 153:15 PM - 4:00 PM CST
      Room 201DEF

    , AWS

    , Operation and Maintenance Manager, 幣託集團 BitoGroup

    Session Topic: AI/ML, Architecture, Cloud Operations, Kubernetes
    Session Type: Breakout session
    Level: 300 – Advanced
    Session Features: Customer story
  • Generative AI
    Innovation & Transformation
    Automation
    Agentic AI
    DevOps
    Games
    Business Executive
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Kiro
    Breakout session
    Session Catalog
    300 – Advanced
    AI/ML
    Architecture
    DevOps & Developer Productivity
    3:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    本場次分享鈊象電子如何運用 Kiro CLI 的多代理架構，在 AWS 上為遊戲開發打造專業分工的 AI 代理團隊。展示 Server 與 Client 開發團隊各自配置專屬 Agent，並透過協調 Agent 實現跨職能的 agent-to-agent 串接與資訊確認。結合 multi-Kiro 工作流加速團隊協作。並分享 prompt 工程與 context 管理的實戰心法，讓 AI 包辦繁瑣細節，開發者專注於創意本身。(See how IGS leverages Kiro CLI's multi-agent architecture to build specialized AI agent teams that accelerate game development on AWS. We demonstrate how Server and Client development teams each operate dedicated agents, with a coordinator agent bridging cross-functional communication through agent-to-agent orchestration. Using multi-Kiro workflows, IGS streamlines collaboration across teams — all driven by natural terminal conversation. Plus practical tips on prompt engineering and context management so AI handles the tedious work while developers stay focused on creativity.)
    • Wednesday, Jul 153:15 PM - 4:00 PM CST
      3F 大會堂

    , Solutions Architect, AWS

    , 主任工程師, 鈊象電子股份有限公司（International Games System Co., Ltd.）

    Session Topic: AI/ML, Architecture, DevOps & Developer Productivity
    Session Type: Breakout session
    Level: 300 – Advanced
  • Lightning talk
    Session Catalog
    3:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    AI 正在重塑世界——但這對於建構和維運其背後基礎設施的人來說，意味著什麼？在這場 25 分鐘的演講中，您將聽到 AWS 內部的兩個視角：一位資料中心經理分享 AI 工作負載如何改變第一線的營運模式，以及一位業務開發經理談論激增的客戶需求如何創造前所未有的職涯機會。無論您偏向技術或商業領域，這場分享都將帶您一窺科技界成長最快的領域之一——以及為什麼現在正是加入的最佳時機。
    • Wednesday, Jul 153:35 PM - 4:00 PM CST
      2F Developer Community Zone

    , BD Lead, FSI & Retail, AWS

    , DCO cluster manager, AWS

    Session Type: Lightning talk

4:00 PM CST - Wednesday, Jul 15

  • Containers
    DevOps
    Developer / Engineer
    DevOps Engineer
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Amazon Elastic Kubernetes Service (Amazon EKS)
    Breakout session
    Session Catalog
    300 – Advanced
    Architecture
    Cloud Operations
    Kubernetes
    Serverless Compute & Containers
    4:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    當 AI 工作負載對計算資源、節點健康與快速擴縮的需求與日俱增，Amazon EKS Auto Mode 將基礎設施佈建、節點修補與動態擴縮的管理責任移交給 AWS，讓平台工程師得以將精力從維運日常轉移到 AI 應用的開發與交付。本場次透過共同責任模型的前後對比、實機示範與企業客戶案例，說明如何在不增加維運人力的情況下，打造可靠的 AI 工作負載執行平台。(As AI workloads place growing demands on compute resources, node health, and dynamic scaling, Amazon EKS Auto Mode shifts the responsibility for infrastructure provisioning, node patching, and autoscaling to AWS — freeing platform engineers to focus on AI application development and delivery rather than day-to-day operations. Through a before-and-after comparison of the shared responsibility model, a live demo, and real customer examples, this session shows how to build a reliable platform for AI workloads without increasing operational headcount.)
    • Wednesday, Jul 154:15 PM - 5:00 PM CST
      Room 201DEF

    , Cloud Support Engineer II, Amazon Web Services

    Session Topic: Architecture, Cloud Operations, Kubernetes, Serverless Compute & Containers
    Session Type: Breakout session
    Level: 300 – Advanced
  • Generative AI
    Innovation & Transformation
    Automation
    Agentic AI
    Advertising & Marketing
    Advisor / Consultant
    Business Executive
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Sales / Marketing
    Solution / Systems Architect
    Tech Explorer
    Amazon Bedrock
    Breakout session
    Customer story
    Session Catalog
    300 – Advanced
    AI/ML
    Architecture
    Business Applications
    4:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    讓購物聊天機器人進化為自主導購 Agent，不只靠 prompt engineering，還需要結構化的工具編排、有狀態的工作流程，以及生產級基礎設施。本場次展示 TapPay 如何用 Strands Agents 實現 Agent 邏輯，並以 Amazon Bedrock AgentCore 達成可擴展的安全部署。(Turning a shopping chatbot into an autonomous sales agent requires more than prompt engineering, but it demands structured tool orchestration, stateful workflows, and production-grade infrastructure. This session shows how TapPay achieved this with Strands Agents for agent logic and Amazon Bedrock AgentCore for scalable, secure deployment.)
    • Wednesday, Jul 154:15 PM - 5:00 PM CST
      3F 大會堂

    , Senior Solutions Architect, AWS

    , Vice President, 喬睿科技股份有限公司 Cherri Tech

    Session Topic: AI/ML, Architecture, Business Applications
    Session Type: Breakout session
    Level: 300 – Advanced
    Session Features: Customer story
  • Agentic AI
    Developer / Engineer
    DevOps Engineer
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Kiro
    Breakout session
    Session Catalog
    200 – Intermediate
    AI/ML
    DevOps & Developer Productivity
    4:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    瞭解 AI 代理人如何透過動態載入專屬能力，即時化身領域專家，全面加速全端開發與部署流程。聆聽我們的專家與合作夥伴分享第一手見解，並親眼見證 Kiro Powers 的實戰演示。(Join this breakout session to learn how AI agents turn into experts instantly by dynamically loading powers to accelerate full-stack development and deployment. Hear from our expert and partners and see demos of powers in action.)
    • Wednesday, Jul 154:15 PM - 5:00 PM CST
      Room 201ABC

    , Sr. Cloud Support Engineer, AWS

    Session Topic: AI/ML, DevOps & Developer Productivity
    Session Type: Breakout session
    Level: 200 – Intermediate
  • Global Infrastructure
    Containers
    Kubernetes
    Cloud Security Specialist
    Developer / Engineer
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Amazon Elastic Kubernetes Service (Amazon EKS)
    Breakout session
    Session Catalog
    300 – Advanced
    Architecture
    Cloud Operations
    Kubernetes
    4:00 p.m. Wednesday, Jul 15
    Wednesday, Jul 15
    傳統輪詢式負載均衡無法應對 LLM 推論請求成本高度不均的特性。本場次介紹 ALB Target Optimizer 的拉取式模型如何讓目標端主動回報容量狀態，在過載發生前即時因應。並深入探討在 Amazon EKS 上搭配 GPU 推論端點部署 Target Optimizer 的架構模式。(Discover how ALB Target Optimizer's pull-based model transforms load balancing for AI/ML inference. Learn why traditional round-robin fails for LLM serving, how targets actively report capacity to eliminate overload, and architecture patterns for deploying Target Optimizer with GPU-based inference endpoints on Amazon EKS. )
    • Wednesday, Jul 154:15 PM - 5:00 PM CST
      Room 101BC

    , Cloud Support Engineer, AWS

    Session Topic: Architecture, Cloud Operations, Kubernetes
    Session Type: Breakout session
    Level: 300 – Advanced

9:00 AM CST - Thursday, Jul 16

  • Artificial Intelligence
    Cloud Operations
    Monitoring & Observability
    Agentic AI
    DevOps
    Developer / Engineer
    DevOps Engineer
    Solution / Systems Architect
    AWS DevOps Agent
    Workshop
    300 – Advanced
    Hands-on
    Laptop required
    Session Catalog
    Apply to attend
    9:00 a.m. Thursday, Jul 16
    Thursday, Jul 16
    在現代雲端架構中，系統故障排除往往是工程師最耗時且壓力最大的工作。凌晨兩點的告警、跨服務的級聯故障、缺乏上下文的根因分析——這些場景每天都在發生。AWS DevOps Agent 是 AWS 推出的全方位 AI 運維代理，能在事件發生的第一時間自動啟動調查、快速定位 RootCause、提供修復建議，並在事後持續分析模式以主動預防未來同類故障。無論您的架構是 Serverless、容器化微服務、數據管線還是 AI/ML 工作負載，DevOps Agent 都能顯著縮短平均修復時間，本工作坊將帶您親手操作實戰，從故障注入到 AI 驅動的自動根因分析，完整體驗 DevOps Agent 如何改變您的維運方式。
    • Thursday, Jul 169:30 AM - 12:00 PM CST
      Room 103

    , Sr. Cloud Support Engineer, AWS

    Session Topic: Artificial Intelligence, Cloud Operations
    Session Type: Workshop
    Session Features: Hands-on, Laptop required

10:00 AM CST - Thursday, Jul 16

  • Keynote
    100 – Foundational
    Customer story
    Session Catalog
    Thursday, Jul 16
    10:00 a.m. Thursday, Jul 16
    當 AI 代理（AI Agent）從技術前沿走入企業核心，真正的變革才正要開始。在 Day 2 Keynote 中，AWS 將聚焦產業策略與代理式 AI（Agentic AI）的商業價值，分享 AWS 在台灣的最新佈局與產業深耕方向。你將聽到來自全球領先企業的高階主管，親自講述他們如何以 AI 驅動業務轉型、創造可衡量的商業成果。無論你是正在評估 AI 策略的企業決策者，還是推動組織創新的業務領導人，這場 Keynote 將為你揭示代理式 AI 如何從願景化為實際的競爭優勢。(As AI Agents move from the technological frontier into the enterprise core, the real transformation is just beginning. In the Day 2 Keynote, AWS will focus on industry strategy and the business value of Agentic AI, sharing AWS's latest initiatives and industry deepening directions in Taiwan. You will hear senior executives from globally leading enterprises personally recount how they are driving business transformation with AI and delivering measurable business outcomes. Whether you are an enterprise decision-maker evaluating AI strategy, or a business leader driving organizational innovation, this Keynote will reveal how Agentic AI transforms from vision into tangible competitive advantage.)
    • Thursday, Jul 1610:00 AM - 11:00 AM CST
      3F 大會堂

    , Managing Director, Taiwan, Amazon Web Services, Inc.

    , Sr Manager, Solutions Architecture, Amazon Web Services, Inc.

    , APAC Head of Applied AI, Anthropic 

    , Calypso Senior Director APAC, Nasdaq

    , Deputy Director, Korean Broadcasting System (KBS)

    Session Type: Keynote
    Session Features: Customer story

11:00 AM CST - Thursday, Jul 16

  • Application Security
    Automation
    Culture of Security
    Agentic AI
    DevOps
    Cloud Security Specialist
    Developer / Engineer
    DevOps Engineer
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    AWS Security Agent
    Breakout session
    Customer story
    Session Catalog
    300 – Advanced
    AI/ML
    Architecture
    Security, Compliance, & Identity
    Thursday, Jul 16
    11:00 a.m. Thursday, Jul 16
    本場次將解析 Agentic AI 的新興威脅，並介紹 AWS Security Agent 如何透過 AI 自動化安全審查與主動式防護強化應用安全；同時分享 CyCraft XecGuard 如何結合 AWS Guardrail API，提供 Prompt Attack 防護、風險偵測與 AI 安全治理能力，協助企業建立可控、可管、可驗證的 Agentic AI 安全防護架構。(This session will analyze the emerging threats of Agentic AI and introduce how AWS Security Agent automates security reviews and strengthens application security through proactive protection powered by AI. Additionally, we will share how CyCraft XecGuard integrates with the AWS Guardrail API to deliver Prompt Attack protection, risk detection, and AI security governance capabilities — helping enterprises establish a controllable, manageable, and verifiable Agentic AI security protection architecture.)
    • Thursday, Jul 1611:30 AM - 12:15 PM CST
      Room 101BC

    , Lead Solutions Architecture, AWS

    , Co-founder and CTO, CyCraft 奧義智慧

    Session Topic: AI/ML, Architecture, Security, Compliance, & Identity
    Session Type: Breakout session
    Level: 300 – Advanced
    Session Features: Customer story
  • Global Infrastructure
    Generative AI
    Management & Governance
    Well-Architected Framework
    Automation
    Agentic AI
    Manufacturing & Industrial
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Breakout session
    Session Catalog
    300 – Advanced
    AI/ML
    Analytics
    Architecture
    Business Applications
    Cloud Operations
    Thursday, Jul 16
    11:00 a.m. Thursday, Jul 16
    一家全球製造企業歷經多年漸進式轉型：400+ 應用上雲、整合跨 100 座工廠的數據孤島、完成 shop floor 系統現代化。本場次拆解三大階段的實務做法，涵蓋大規模遷移的組織變革與決策框架、應用現代化帶來的 67–83% 部署加速與 74% 資料處理時間縮減，以及如何在穩固基礎上一週內落地生成式 AI 產線助手。帶走一套製造業從遷移到 AI 創新的分階段 playbook。(A global manufacturing enterprise underwent years of progressive transformation: migrating 400+ applications to the cloud, consolidating data silos across 100+ factories, and modernizing shop floor systems. This session breaks down the practical approaches across three major phases, covering organizational change management and decision frameworks for large-scale migration, application modernization delivering 67–83% deployment acceleration and 74% reduction in data processing time, and how to deploy a generative AI production line assistant within one week on a solid foundation. Walk away with a phased playbook for manufacturing — from migration to AI innovation.)
    • Thursday, Jul 1611:30 AM - 12:15 PM CST
      Room 201DEF

    , Principal Solutions Architect, AWS

    , Solutions Architect, AWS

    Session Topic: AI/ML, Analytics, Architecture, Business Applications, Cloud Operations
    Session Type: Breakout session
    Level: 300 – Advanced
  • Well-Architected Framework
    Automation
    Agentic AI
    Business Executive
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Amazon Aurora
    Breakout session
    Customer story
    Session Catalog
    300 – Advanced
    AI/ML
    Architecture
    Business Applications
    Database
    Storage
    Thursday, Jul 16
    11:00 a.m. Thursday, Jul 16
    AI Agent 正從被動回應演進為主動決策與執行，MCP 提供標準化介面讓 Agent 安全存取外部工具與資料。零售業的商品目錄、庫存、訂單與會員資料多存於關聯式資料庫，將這些 system of record 轉化為 Agent 可操作的工具層，是落地 Agentic Workflow 的關鍵。(AI Agents are evolving from passive responders to proactive decision-makers and executors. MCP provides a standardized interface that enables Agents to securely access external tools and data. In retail, product catalogs, inventory, orders, and membership data predominantly reside in relational databases. Transforming these systems of record into a tool layer that Agents can operate on is the key to implementing Agentic Workflows in production.)
    • Thursday, Jul 1611:30 AM - 12:15 PM CST
      Room 201ABC

    , Sr. Database GTM Specialist, AWS

    , Business Manager, NC AI

    Session Topic: AI/ML, Architecture, Business Applications, Database, Storage
    Session Type: Breakout session
    Level: 300 – Advanced
    Session Features: Customer story
  • Automation
    Agentic AI
    Financial Services
    Business Executive
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Amazon Bedrock
    Amazon OpenSearch Serverless
    Amazon OpenSearch Service
    OpenSearch
    Breakout session
    Customer story
    Session Catalog
    300 – Advanced
    AI/ML
    Architecture
    Business Applications
    Serverless Compute & Containers
    Thursday, Jul 16
    11:00 a.m. Thursday, Jul 16
    虛擬資產公司如何打造可跨部門服務的可擴展 AI Agent 平台？HOYA BIT 分享如何運用 Amazon Bedrock AgentCore 與 AWS 雲端服務建構企業級 AI Agent 平台，透過標準化 Agent Runtime、工具整合與治理機制，實現業務流程自動化、知識存取與決策輔助，加速 AI Agent 應用落地與規模化部署。(How does a digital asset company build a scalable AI Agent platform that serves across departments? HOYA BIT shares how they built an enterprise-grade AI Agent platform with Amazon Bedrock AgentCore and AWS cloud services — enabling business process automation, knowledge access, and decision support through a standardized Agent Runtime, tool integration, and governance framework designed for rapid adoption and scale.)
    • Thursday, Jul 1611:30 AM - 12:15 PM CST
      3F 大會堂

    , Solutions Architect, AWS

    , Founder/Chairperson, HOYA BIT Exchange

    , Head of AI Engineering, HOYA BIT Exchange

    Session Topic: AI/ML, Architecture, Business Applications, Serverless Compute & Containers
    Session Type: Breakout session
    Level: 300 – Advanced
    Session Features: Customer story
  • Generative AI
    Automation
    Agentic AI
    DevOps
    Advertising & Marketing
    Automotive
    Financial Services
    Media & Entertainment
    Travel & Hospitality
    Business Executive
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    Student
    Tech Explorer
    Code talk
    100 – Foundational
    Hands-on
    Laptop required
    Peer networking
    Session Catalog
    Thursday, Jul 16
    11:00 a.m. Thursday, Jul 16
    【當 AI 走進生活的每一個角落，下一個被突破的產業痛點，由你來定義！】 【由 2-5 人組成隊伍進行報名 Summit & Hackathon，選擇欲參與的題目，並於活動內運用 AWS 提供之雲端環境與 AI 工具進行開發，提出最有創意的解決方案。】 【不接受個人報名，請組隊參加！主辦單位將不協助媒合！】 【總獎金高達 45 萬台幣！僅限 100 人參加！先報先得！】 從早晨打開新聞推播、通勤路上聽聽音樂、午餐時看股市庫存、到晚上滑直播 — AI 已經無聲滲透進我們生活的每一處。然而，在這些日常體驗的背後，百工百業仍有許多產業痛點等待被 AI 解決，包含金融投資人的決策焦慮、物流業的效率瓶頸、媒體內容的個人化困境、娛樂產業的互動創新需求。 AWS Summit Taipei 2026 - 百工百業瘋 AI - AI Everywhere Hackathon，集結了台灣 4 大產業的領導企業 - 【CMoney、KKCompany、陽明海運、浪 LIVE】，拋出他們最真實的業務挑戰、提供第一手產業數據，由 AWS 作為 AI 技術後盾，並且邀約你嶄露傑出一手來解題。這不只是場競賽，更是 AI 創新加速器，將由企業提供數據與命題、AWS 提供技術環境，你來提供最有創意的解法。
    • Thursday, Jul 1611:30 AM - 5:30 PM CST
      4F 鳳凰廳
    Session Type: Code talk
    Session Features: Hands-on, Laptop required, Peer networking

1:00 PM CST - Thursday, Jul 16

  • Financial Services
    Business Executive
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Breakout session
    Customer story
    Session Catalog
    Apply to attend
    1:00 p.m. Thursday, Jul 16
    Thursday, Jul 16
    本場議程首先介紹 AWS 上的 Web3 關鍵基礎服務 — 從 Amazon Managed Blockchain 的託管節點存取、Blockchain Node Runners 的彈性自建部署，到 Nitro Enclaves 為加密金鑰與敏感運算提供硬體層級隔離保護，完整涵蓋企業級區塊鏈基礎設施所需的安全性、可擴展性與合規能力。隨後，我們邀請區塊科技股份有限公司（Blockchain Security Corp.）分享其區塊鏈節點上雲的實戰經驗：如何運用 AWS 實現節點的雲端原生部署與彈性擴展，並透過 Nitro Enclaves 建立可信執行環境 — 將原本受限於地端固定容量的節點架構，轉化為可隨鏈上需求動態伸縮的雲原生基礎設施。 (This session begins with an overview of AWS's core Web3 building blocks — from Amazon Managed Blockchain for managed node access and Blockchain Node Runners for flexible self-hosted deployments, to Nitro Enclaves providing hardware-level isolation for cryptographic keys and sensitive workloads — covering the security, scalability, and compliance capabilities required for enterprise blockchain infrastructure. We then invite Blockchain Security Corp. to share their hands-on experience migrating blockchain nodes to the cloud: how they leveraged AWS for cloud-native node deployment and elastic scaling, and used Nitro Enclaves to establish trusted execution environments — transforming a capacity-constrained on-premises node architecture into a dynamically scalable, cloud-native infrastructure that adapts to on-chain demand.)
    • Thursday, Jul 161:30 PM - 2:15 PM CST
      Room 102

    , Business Developer Manager, AWS

    , R&D Director, Blockchain Security Corp.

    Session Type: Breakout session
    Session Features: Customer story
  • Application Security
    Business Executive
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Lightning talk
    Community-led
    Session Catalog
    100 – Foundational
    Security, Compliance, & Identity
    1:00 p.m. Thursday, Jul 16
    Thursday, Jul 16
    在 AI 與雲端全面落地的時代，真正的資安挑戰已從「資料儲存與傳輸」延伸到「資料使用中」。當敏感資料進入記憶體處理時，如何確保即使系統管理員或雲端平台也無法存取？這正是機密運算（Confidential Computing）要解決的關鍵問題。 本次演講將從資料三態的安全風險出發，解析機密運算的核心概念與可信任執行環境（TEE）運作原理，並以 AWS Nitro Enclaves 為實務主軸，說明如何建立硬體隔離的安全運算環境，保護 LLM 推論與敏感資料處理。
    • Thursday, Jul 161:30 PM - 1:55 PM CST
      2F Developer Community Zone

    , AWS Security Hero, AWS Security Hero

    Session Topic: Security, Compliance, & Identity
    Session Type: Lightning talk
    Level: 100 – Foundational
    Session Features: Community-led
  • Artificial Intelligence
    Developer Tools
    Generative AI
    Agentic AI
    DevOps
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    Tech Explorer
    Amazon Bedrock
    Kiro
    Workshop
    300 – Advanced
    Hands-on
    Laptop required
    Session Catalog
    Apply to attend
    1:00 p.m. Thursday, Jul 16
    Thursday, Jul 16
    AI 正在重新定義軟體開發與維運的方式。本工作坊將帶您深入體驗 Kiro CLI — AWS 推出的終端機原生 AI 開發工具，結合 Steering（行為準則）、Skills（Agent 技能） 與 MCP Servers（外部工具整合） 三大核心機制，實現對 Kubernetes 叢集的 AI 驅動式維運。在本工作坊中，您將：認識 Kiro CLI 核心架構：了解 Steering files 如何定義 agent 行為規範、Skills 如何封裝可重複執行的維運技能、MCP 如何讓 AI agent 即時連接外部工具與資料安裝與設定 MCP Servers：將 Amazon EKS MCP Server 連接至 Kiro CLI，讓 AI agent 具備叢集感知能力 — 用自然語言查詢 Pod 狀態、部署應用、診斷故障 Hands-on Lab — AI-Powered Kubernetes Ops on EKS Auto Mode：從叢集建立、應用部署、自動擴展，到故障排除，全程在終端機中以 Kiro CLI 對話完成快速了解 Claude Code 在 AWS 上的選項（10–15 分鐘）：Claude on Amazon Bedrock、Claude Platform on AWS、Claude for Enterprise 三種企業部署方案
    • Thursday, Jul 161:30 PM - 4:30 PM CST
      Room 103

    , Solutions Architect, AWS

    , Sr. Cloud Support Engineer, AWS

    Session Topic: Artificial Intelligence, Developer Tools
    Session Type: Workshop
    Session Features: Hands-on, Laptop required
  • Cloud Security Specialist
    Developer / Engineer
    DevOps Engineer
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Amazon GameLift
    Breakout session
    Session Catalog
    300 – Advanced
    1:00 p.m. Thursday, Jul 16
    Thursday, Jul 16
    Palworld 提供官方專屬伺服器（Official Dedicated Server），讓新玩家無需自行架設即可體驗多人連線。著眼於長期營運，我們將這些伺服器從 Kubernetes 遷移至 Amazon GameLift。GameLift 的設計以 Session-based 遊戲託管為核心，而 Palworld 則為每台伺服器保留獨立的存檔資料。本場次將分享我們如何處理這項架構差異、部署流程的變化，以及相較於原先 Kubernetes 營運方式所獲得的優勢與取捨。
    • Thursday, Jul 161:30 PM - 2:15 PM CST
      Room 201DEF

    , SA, AWS

    , Platform Engineer, Pocketpair

    Session Type: Breakout session
    Level: 300 – Advanced
  • Data Protection
    Well-Architected Framework
    Automation
    DevOps
    Financial Services
    Amazon Simple Storage Service (Amazon S3)
    Breakout session
    Session Catalog
    300 – Advanced
    Architecture
    Business Applications
    Database
    Storage
    1:00 p.m. Thursday, Jul 16
    Thursday, Jul 16
    MaiCoin 的加密貨幣交易所 24/7 全天候運轉，為在不干擾來源系統的前提下建立分析層，設計了 Unified Data Platform：透過 CDC 將營運資料庫變更即時串流至 Amazon S3 上的 Apache Iceberg tables，形成與 system of record 同步的開放格式分析底座。本場次分享整體架構、CDC pipeline 選型、24/7 工作負載下的 schema 演進策略，以及在交易所規模下維運 Iceberg 資料平台的實務經驗。
    • Thursday, Jul 161:30 PM - 2:15 PM CST
      Room 201ABC

    , Sr. Database Specialist Solutions Architect, AWS

    , 資產負債管理部門 - 業務襄理, 現代財富科技股份有限公司

    Session Topic: Architecture, Business Applications, Database, Storage
    Session Type: Breakout session
    Level: 300 – Advanced
  • Application Security
    Culture of Security
    Agentic AI
    Identity & Access Management
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Amazon Cognito
    AWS Identity and Access Management (IAM)
    AWS Security Agent
    Breakout session
    Customer story
    Session Catalog
    200 – Intermediate
    Architecture
    Security, Compliance, & Identity
    1:00 p.m. Thursday, Jul 16
    Thursday, Jul 16
    當企業開始在 AWS 上部署 AI Agent 驅動核心業務流程，一個關鍵問題隨之浮現：誰來確保每個 Agent 只做它被授權的事、每個租戶的資料不互相滲透、每一筆操作都留下符合稽核要求的紀錄？本場議程以 HeartBot 企業級 AI Agent 治理平台為例，分享如何在模型無關的架構上建立一層完整的安全與治理框架：透過 Cognito 實現多租戶零信任身份隔離、以 DLP 與 NER 在 Agent 執行期間動態遮罩敏感個資、並為每個租戶建立獨立的 AI interaction audit trail，讓操作紀錄能直接對應 ISO 27001 與 EU AI Act 的稽核要求。最後也將展望如何透過 Cognito Identity Pool 並結合 AWS Kiro和 AWS Security Agent進一步收緊身份邊界、加速合規就緒的開發節奏。如果您正在評估如何讓企業客戶放心把核心業務交給 AI Agent 平台執行，這場演講是為您設計的。(As enterprises deploy AI agents on AWS to drive core business processes, a critical question emerges: who ensures each agent only does what it's authorized to do, that tenant data never bleeds across boundaries, and that every action leaves an audit trail that satisfies compliance requirements? This session draws on HeartBot's enterprise AI agent governance platform to show how a complete security and governance layer can be built on a model-agnostic architecture: zero-trust multi-tenant identity isolation through Cognito, real-time PII masking via DLP and NER during agent execution, and per-tenant AI interaction audit trails designed to map directly to ISO 27001 and EU AI Act requirements. We'll close by looking ahead to Cognito Identity Pool and AWS Kiro, AWS Security Agent as the next layer of identity enforcement and compliance-ready development velocity. If you're evaluating how to give enterprise customers the confidence to hand their core business processes over to an AI agent platform — this session is for you.)
    • Thursday, Jul 161:30 PM - 2:15 PM CST
      Room 101BC

    , Cloud Support Engineer, AWS

    , Founder, Heartbot 聆機科技

    , Cloud Support Engineer, Amazon Web Services

    Session Topic: Architecture, Security, Compliance, & Identity
    Session Type: Breakout session
    Level: 200 – Intermediate
    Session Features: Customer story
  • VMware
    Business Executive
    Developer / Engineer
    DevOps Engineer
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    Tech Explorer
    Breakout session
    Session Catalog
    300 – Advanced
    AI/ML
    Architecture
    Business Applications
    Cloud Operations
    Compute
    Hybrid Cloud
    1:00 p.m. Thursday, Jul 16
    Thursday, Jul 16
    本場次深入拆解 AWS Transform 的伺服器遷移完整生命週期，探索其 AI 代理程式如何自動化遷移歷程，涵蓋本地環境智慧探索、AI 生成的波次計畫與網路架構規劃，並透過生成式 AI 實現基礎設施即程式碼與網路轉換的自動化。透過真實遷移情境演示，帶走可立即上手的 AI 增強遷移技巧，將業務衝擊降至最低。(This session covers the full server migration lifecycle with AWS Transform, showing how AI-powered agents automate discovery, wave planning, and network architecture design. See generative AI in action for infrastructure-as-code generation and network transformation, with real-world migration scenarios demonstrating how to modernize at scale while minimizing business disruption.)
    • Thursday, Jul 161:30 PM - 2:15 PM CST
      3F 大會堂

    , Sr. Solutions Architect, AWS

    , Solutions Architect, AWS

    Session Topic: AI/ML, Architecture, Business Applications, Cloud Operations, Compute, Hybrid Cloud
    Session Type: Breakout session
    Level: 300 – Advanced
  • Agentic AI
    Sales / Marketing
    Kiro
    Lightning talk
    Community-led
    Session Catalog
    100 – Foundational
    AI/ML
    1:00 p.m. Thursday, Jul 16
    Thursday, Jul 16
    在 AI 快速發展的時代，打造專屬 AI 工具已不再是工程師的專利。本場分享將以 HR 的實務場景出發，帶領你了解如何運用 AWS 的生成式 AI 工具 Kiro，快速打造屬於自己的 AI 小幫手。從需求發想、流程設計，到 AI 協助生成內容與自動化任務，HR 也能在無需深厚技術背景的情況下，建立能實際應用於工作中的 AI 解決方案。這不只是一場工具介紹，更是一個讓非技術職能重新掌握生產力的起點。
    • Thursday, Jul 161:55 PM - 2:20 PM CST
      2F Developer Community Zone

    , AWS User Group DevSecOps Taiwan Founder, ifus

    Session Topic: AI/ML
    Session Type: Lightning talk
    Level: 100 – Foundational
    Session Features: Community-led

2:00 PM CST - Thursday, Jul 16

  • Data Protection
    Financial Services
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    System Administrator
    Tech Explorer
    Amazon Aurora
    Lightning talk
    Community-led
    Session Catalog
    300 – Advanced
    Analytics
    Database
    Storage
    Thursday, Jul 16
    2:00 p.m. Thursday, Jul 16
    當您的單表數據突破 800GB、總行數超過 20 億時，資料庫架構的每一毫秒延遲都決定了業務的成敗。這次我們將深入 Aurora 的底層分散式特性，並分享一套在極端高壓環境下穩定運行的實戰架構。 我們只談實戰。深入解析如何透過 CLI 精準配置 Custom Endpoints，並靈活運用 Serverless v2 隔離地獄級查詢壓力。 帶走這份 「Aurora 效能優化檢查清單」 ，讓您的數據架構不僅僅是 「堪用」 ，而是具備應對未來百倍增長的 「極致彈性」 。(When your single table data exceeds 800GB with over 2 billion rows, every millisecond of latency in your database architecture determines business success or failure. This time, we will dive deep into Aurora's underlying distributed characteristics and share a battle-tested architecture that runs stably under extreme high-pressure environments. We only talk real-world practice. We will provide an in-depth analysis of how to precisely configure Custom Endpoints via CLI, and flexibly leverage Serverless v2 to isolate nightmare-level query pressure. Take away this "Aurora Performance Optimization Checklist" so that your data architecture is not merely "adequate," but possesses the "ultimate elasticity" to handle hundredfold future growth.)
    • Thursday, Jul 162:20 PM - 2:45 PM CST
      2F Developer Community Zone

    , 資料庫工程師 DBA, GTS Platfrom Services Limited

    Session Topic: Analytics, Database, Storage
    Session Type: Lightning talk
    Level: 300 – Advanced
    Session Features: Community-led
  • Agentic AI
    Financial Services
    Business Executive
    Cloud Security Specialist
    Data Engineer
    Data Scientist
    Developer / Engineer
    DevOps Engineer
    Entrepreneur (Founder/Co-Founder)
    IT Administrator
    IT Executive
    IT Professional / Technical Manager
    Solution / Systems Architect
    Tech Explorer
    Amazon Bedrock
    Breakout session
    Customer story
    Session Catalog
    AI/ML
    Apply to attend
    Thursday, Jul 16
    2:00 p.m. Thursday, Jul 16
    本場議程探討一家全球證券金融科技業者如何將其即時市場數據平台從地端遷移至 AWS — 實現低於 100 毫秒的全球延遲，並將新區域上線時間從數週縮短至分鐘級。在此雲原生數據基礎之上，我們展示如何利用 Amazon Bedrock 上的Multi-Agents AI 重新定義量化回測：智能體自主獲取即時行情、生成交易策略、執行回測並評估結果 — 將原本需要大量手動編碼的流程轉化為對話式、可迭代的智能體驗。 (This session explores how a global securities fintech provider migrated its real-time market data platform from on-premises to AWS — achieving sub-100ms global latency and reducing new region deployment from weeks to minutes. Building on this cloud-native data foundation, we demonstrate how multi-agent AI on Amazon Bedrock reinvents quantitative backtesting: agents autonomously fetch live quotes, generate trading strategies, execute backtests, and evaluate results — transforming a manual, code-heavy workflow into a conversational, iterative experience.)
    • Thursday, Jul 162:25 PM - 3:05 PM CST
      Room 102

    , Sr. GenAI Solution Architect, AWS

    , Solution Architect, AWS

    , President, ViewTrade Technology

    Session Topic: AI/ML
    Session Type: Breakout session
    Session Features: Customer story