← Financial Cloud Cloud Cloud Club · AWS Re:cap

Vertex Macro | Financial Cloud Cloud · AWS Re:cap

AWS Re:cap 06: Blog to Slides in Three Minutes with Kiro

Speaker: Haowen Huang

Session: 06

Session
Summit Dev Lounge2026 Re:cap
01 Encode Architecture as Steering for AI Agents
Summit Dev Lounge2026 Re:cap
02 Agent Harness Is the Real Engineering Moat
Summit Dev Lounge2026 Re:cap
03 Ask Observability Data in Plain Language
Summit Dev Lounge2026 Re:cap
04 Serverless AR Game with Bedrock AgentCore
Summit Dev Lounge2026 Re:cap
05 Multi-Agent Quant Backtesting on AgentCore
Summit Dev Lounge2026 Re:cap
06 Blog to Slides in Three Minutes with Kiro
Summit Dev Lounge2026 Re:cap
AWS Community Day Hong Kong 2025 Re:cap
02 AWS Compliance with Terraform
AWS Community Day Hong Kong 2025 Re:cap
03 Beginner to Builder An AWeSome Cloud Journey
AWS Community Day Hong Kong 2025 Re:cap
04 Team-First Serverless Engineering with Laravel & Bref
AWS Community Day Hong Kong 2025 Re:cap
05 Event Opening - AWS Community Day Hong Kong 2025
AWS Community Day Hong Kong 2025 Re:cap
06 Agent-to-Agent: Building Interoperable AI on AWS
AWS Community Day Hong Kong 2025 Re:cap
07 Utilize another telemetry data for faster improvement with AI agent
AWS Community Day Hong Kong 2025 Re:cap
08 Graduating from Vibe Coding: Spec-Driven Development with Kiro
AWS Community Day Hong Kong 2025 Re:cap
09 Automated Testing using MCP & AI Agents
AWS Community Day Hong Kong 2025 Re:cap
10 Modernizing Telecom Security ML Powered Approach
AWS Community Day Hong Kong 2025 Re:cap
11 Rethinking GenAI Agent: RAG & MCP
AWS Community Day Hong Kong 2025 Re:cap
12 Disaster and Emergency Response with TAK and AWS
AWS Community Day Hong Kong 2025 Re:cap
13 Rethinking Serverless Application Workflows from a Testing Perspective
AWS Community Day Hong Kong 2025 Re:cap
14 Practical AWS FinOps for Cloud Success
AWS Community Day Hong Kong 2025 Re:cap
15 AI-Powered Global Pure-Alpha Macro Trades on AWS: Revolutionizing Risk-Adjusted Asset Returns
AWS Community Day Hong Kong 2025 Re:cap
FSI Recap
01 Modern Trade Lifecycle: Trading to Settlement
FSI Recap
02 Goldman Sachs: Fast Track your applications onto Cloud - AWS Re:cap Q1/2023
FSI Recap
03 Zurich Insurance Group: Building an Effective Log Management Solution on AWS
FSI Recap
04 FSI Meetup 2025 Q4 - Brex Database Disaster Recovery
FSI Recap
05 FSI Meetup 2025 Q4 - A Graviton Migration Success Story
FSI Recap
06 FSI Meetup 2025 Q4 - Stifel Modern Data Platform
FSI Recap
07 FSI Meetup 2025 Q4 - Financial Transaction Data Reconciler PayPal
FSI Recap
08 FSI Meetup 2025 Q4 - Scaling Resilience
FSI Recap
09 Maximizing AI Inference Cost Efficiency: Strategic Adoption of AWS GPU Instances
FSI Recap
10 Advanced Agentic AI Design Patterns
FSI Recap
11 Build New Modern Apps on AWS
FSI Recap
AWS re:Invent 2025
01 Coinbase re:Invent Recap (IND3312)
AWS re:Invent 2025
02 Building the Future Trading Platform Leveraging AI and AWS
AWS re:Invent 2025
03 Trading Innovation: Jefferies' AI Assistant on Amazon Bedrock (IND3315)
AWS re:Invent 2025
04 How FSI Revolutionized HFT Analytics with Agentic AI (GBL302)
AWS re:Invent 2025
05 Improving Distributed Systems with Amazon Time Sync Featuring Nasdaq
AWS re:Invent 2025
06 Amazon Aurora HA and DR Design Patterns for Global Resilience (DAT442)
AWS re:Invent 2025
07 Building Agentic AI: Amazon Nova Act and Strands Agents in Practice (DEV327)
AWS re:Invent 2025
08 Deep Dive into Amazon Aurora and Its Innovations (DAT441)
AWS re:Invent 2025
09 Deep dive on Amazon S3 (STG407)
AWS re:Invent 2025
10 Nasdaq: Build Resilient Infrastructure for Global Financial Services (HMC327)
AWS re:Invent 2025
11 What's New with AWS Lambda (CNS376)
AWS re:Invent 2025
12 Spec-Driven Development with Kiro (DEV314)
AWS re:Invent 2025
13 Amazon's finops: Cloud cost lessons from a global e-commerce giant (AMZ308)
AWS re:Invent 2025
14 Tick to trade latency trading platforms on aws
AWS re:Invent 2025
Government data
01 The AI Era: The Boundary Between Development and Design Is Disappearing
Government data
02 On-Device Multimodal AI and Smart-City Practice
Government data
03 Large-Model Capability Evaluation and a Method for Landing AI Projects
Government data
04 Controlled End-to-End Automation of Government Development with Cloud Agents
Government data
05 A New Software Ecosystem for the Agent Era, Seen Through Multi-Agent Systems
Government data
06 AI-Driven Macro Quantitative Research and Smart Governance
Government data
07 Authorized Operation of Public Data and Smart-Government Practice
Government data
08 Putting Data Assetization into Practice: Rights, Compliance, Engineering Governance, and Digital-Government Cases
Government data
09 AI for Mental-Health Public Welfare: Governance, Architecture, and Practice of a Trustworthy Platform
Government data
Amarathon 2025 Recap
01 A Developer’s Roadmap to Architecting for Agents
Donnie Prakoso
02 Amazon Bedrock Data Automation
Hafiz Syed Ashir Hassan
03 Multi-Agent on AgentCore
Tan Xin
04 Building Agentic AI Nova Act and Strands Agents in Practice
Haowen Huang
04 Accelerating Migration Projects with Kiro using Spec-Driven Development
Sanchit Dilip Jain
06 From Matching to Understanding: Personalized AI Search Practice Driven by AgentCore Memory
Liu Cao
07 Observe to Optimize – LLM Observability to AIOps Turning real-time insights into intelligent automation
Jimmy Soh
08 Deploying TEAM and Building the Best Engineering Team
Yuji Oshima
09 Five Hard Lessons from Five Years of So-Called Serverless Databases
Renato Losio
14 What if AI does my job How Q Developer CLI and Kiro have changed my daily routine
Miguel Angel Muñoz
16 Velocity with Vigilance: Security Essentials for Amazon Bedrock Agent Development
Brian Tarbox
26 Run OSS LLMs on a Single H100 Smarter, Cheaper, Faster
Adit Modi
28 A Modern Unified Metadata Architecture: New Approaches to Breaking Down Data Silos
Shaofeng Shi
29 Serverless MediaOps: Automating Video Workflows with AI on Amazon Web Services
Luis Valdivia
30 Architecting for Efficiency and Reliability with Performance Testing at Scale
Luis Guirigay
31 Connecting the World Through Open Source: Practical Journey of Technology, Community and Global Developer Relations
Richard Lin
33 Building Streaming Iceberg Tables for Real-Time Logistics Analytics
Fahad Shah
34 Accelerating Large-Scale Robot Strategy Training: An Automated Closed-Loop Architecture Based on Kiro, Trainium, and EKS
Junjie Tang
35 From Vibe to Viable with spec driven development
Ricardo Sueiras
36 Making Cloud Cost Analysis Smarter: Building FinOps Intelligent Agents with Strands and AgentCore
Xiaofei Li
37 Transform Conversational Agentic AIOps for K8s Using CNCF Kagent, K8sGPT, and Nova Sonic
Shaoyi Li

AWS Hong Kong Summit 2026 · Developer Lounge Recap · Article 6 of 6

AWS Magazine · Field Report from Victoria Harbour


Opening: High-Tech Luxury Meets the Waterfront Vibe

If you have ever stood on the floor-to-ceiling glass concourse at the Hong Kong Convention and Exhibition Centre (HKCEC) during the AWS Hong Kong Summit 2026, you know the exact feeling: floor-to-ceiling panoramic views of Victoria Harbour, neon reflections dancing across dark water, and the distinct hum of high-stakes technology meeting global finance.

In the heart of Wan Chai, millions in venture capital and startup deals were being negotiated over hand-poured artisanal tea and espresso. International speakers and cloud architects flew in from Tokyo, London, and Silicon Valley to gather at the AWS Developer Lounge. The environment was electric—a juxtaposition of Hong Kong's signature luxury and AWS’s cutting-edge cloud aesthetic.

Between private investor meetups, startup pitch decks, and technical deep-dives, the Developer Lounge hosted its final session. And as any seasoned tech writer will tell you, every iconic summit ends with a talk that feels like a gift bag you can actually take home and use on your flights.

Closing out the Developer Lounge arc, Haowen Huang delivered a masterclass on turning long-form technical content into polished, executive-ready presentations in three minutes flat.

The Elevator Pitch: Give Kiro an Amazon Web Services blog URL and, in roughly three minutes, receive a custom-themed, production-grade slide deck—complete with Vue components, responsive architecture diagrams, and zero manual layout hassle—ready to export to PowerPoint (PPTX), PDF, or PNG.

+-----------------------------------------------------------------------------------+
THE 3-MINUTE AGENT LOOP

1. [ Tech Blog / URL ]
2. [ Kiro AI Agent Engine ]  -->  reread .kiro/steering/skill.md
3. [ Slidev Engine + theme-aws-dark ]  -->  render Vue components
4. [ Local Preview :3030 ]  -->  export PPTX / PDF / PNG
+-----------------------------------------------------------------------------------+

At AWS Hong Kong Summit 2026, Haowen Huang closed the Developer Lounge with a Kiro and Slidev workflow that turns AWS blog URLs into polished decks in about three minutes. Guided by .kiro/steering/skill.md, the agent fetches content, detects language, applies theme-aws-dark Vue components, and exports PPTX, PDF, or PNG—replacing unpaid meetup design work with a deterministic, version-controllable Markdown pipeline that builders can clone, iterate, and run on the flight home.


Technical Resources & Repositories

Before diving into the hands-on breakdown, bookmark these essential repositories and tools:

● Project GitHub Repository: github.com/hanyun2019/kiro-slidev-ppt-maker

● Slidev Framework: sli.dev

● Kiro Agentic IDE: kiro.dev


The Expensive Hour Nobody Budgets (And How HK Meetups Fix It)

We have all been there. You read a stellar architecture breakdown on Amazon EC2, Amazon Bedrock, or Serverless pipelines, and you want to present it to your engineering team, a local AWS User Group, or potential seed investors.

The traditional slide creation flow is notoriously tedious:

1. Open PowerPoint, Keynote, or Google Slides.

2. Copy and paste paragraphs, painstakingly reducing them to bullet points.

3. Spend an hour alignment-hunting, tweaking font sizes, and adjusting box dimensions.

4. Waste 1 to 2 hours of pure design overhead—even though the core technical content was already written!

In Hong Kong’s fast-moving startup ecosystem, community organizers burn late nights formatting slides before every local meetup. Haowen's project converts that unpaid design labor into a deterministic, three-minute AI agent loop. That represents saved time and resources that can immediately flow back into hiring engineers, securing sponsors, and funding private founder meetups.


The Stack: What Inside the Project Architecture?

This workflow combines Slidev (a developer-first, Markdown-driven slide presentation tool built on Vue 3 and Vite) with Kiro (an AI agentic development environment guided by specs and steering files).

Here is how the project components are structured:

Component Directory / File Technical Role & Purpose
theme-aws-dark/ Custom Slidev theme featuring dark gradient backgrounds inspired by AWS keynotes and HK harbor-night aesthetics. Includes three core Vue components: StatCard, Timeline, and ComparisonTable.
ppt-aws-theme-demo/ & ppt-aws-theme-demo-cn/ English and Chinese reference layout templates demonstrating proper syntax, component usage, and structural standards.
.kiro/steering/skill.md The "Steering Brain" file that enforces directory structures, language auto-detection, density limits, component selection, and output conventions for Kiro.
Preset Mermaid Architecture Snippets Pre-configured Mermaid diagram blocks for popular AWS cloud patterns (Web App, Serverless API, AI Agent, Data Pipeline).

Step-by-Step Hands-On Workflow

Here is how to take any technical blog post from a browser tab to a fully rendered presentation deck in five simple steps.

+-----------------------------------------------------------------------------------+
FIVE-STEP DEVELOPER FLOW

1. Send blog URL to Kiro agent
2. Agent runs automated fetch, language detection, and rendering
3. Launch local hot-reload preview server (localhost:3030)
4. Natural-language iteration and formatting tweaks
5. Export to production PPTX / PDF / PNG
+-----------------------------------------------------------------------------------+

Step 1: Send the Link to Kiro

Open Kiro in your project workspace and issue a plain-language prompt:

Turn this blog post into a PPT:
https://builder.aws.com/content/3EcuvO5GFJY06xS7DyCvk6KQMKk/

If you have specific presentation constraints, add them directly to the instruction:

Turn this blog post into a PPT with no more than 20 slides:
https://builder.aws.com/content/3EcuvO5GFJY06xS7DyCvk6KQMKk/

Step 2: What Kiro Does Under the Hood

When Kiro receives your link, it executes a multi-stage pipeline:

1. Deterministic Web Fetching: First, it attempts a fast HTTP fetch. If the target is a heavy Single Page Application (SPA)—such as builder.aws.com, Substack, or Medium—it automatically switches to a headless browser rendered mode to capture full dynamic content.

2. Language Detection: It analyzes the text density. If the content is primarily Chinese, it configures lang: zh; otherwise, it defaults to lang: en. System typography and Mermaid label fonts adapt automatically.

3. Template Ingestion: Before writing a single byte, Kiro reads ppt-aws-theme-demo/slides.md and all sub-pages under pages/ to absorb exact layout rules and Vue component parameters.

4. Isolated Project Creation: It generates a isolated folder using the naming scheme ppt-{normalized-topic-name}/. Reference templates are never modified, and workspace root hygiene is maintained.

5. Slide Content Generation: It drafts a cover page, section dividers, structured body slides, and a closing page. Decks longer than 10 slides are automatically split into modular chapter files under pages/.

6. Smart Vue Component Assignment: * Single impactful metrics -> StatCard * Sequential events or roadmaps -> Timeline * Side-by-side trade-offs -> ComparisonTable * Complex tabular datasets (>4 metrics) -> Standard Markdown tables

Step 3: Local Hot-Reload Preview

Once generated, Kiro provides a terminal execution command:

npm install && npx slidev ppt-{topic-name}/slides.md

Executing this boots up a local Vite development server at http://localhost:3030/.

● Any manual markdown edits instantly trigger a browser hot-reload.

● Navigate to http://localhost:3030/presenter/ for dedicated presenter mode with speaker notes.

● Navigate to http://localhost:3030/overview/ for a bird's-eye slide grid.

Step 4: Natural Language Iteration

Don't like a specific slide density or background accent? Chat directly with Kiro to adjust the output:

The cost breakdown page has too much text — split it into two slides.
Use a forest-green gradient for the final closing slide.
Replace the bar chart with a Timeline component listing three major milestones.

Kiro edits the corresponding Markdown files instantly while your local server hot-reloads the changes.

Step 5: Export to PowerPoint or PDF

To generate native files for client pitch decks or enterprise meetings, run the export tool:

# Export as editable PowerPoint
npx slidev export ppt-{topic-name}/slides.md --format pptx

# Export as PDF document
npx slidev export ppt-{topic-name}/slides.md

(Note: The initial export will prompt you to install playwright-chromium. Approve it once, and it will be cached for all future runs).


Command Cheat Sheet & Quick-Start Loop

For developers in a hurry, here is the quick command sequence:

Target Goal Command / Action
Generate Deck Paste the blog URL into the Kiro chat prompt.
Local Web Preview npx slidev ppt-{topic}/slides.md
Export PPTX npx slidev export ppt-{topic}/slides.md --format pptx
Export PDF npx slidev export ppt-{topic}/slides.md
Export PNG Images npx slidev export ppt-{topic}/slides.md --format png

Real-World Case Study: Deploying OpenClaw on Amazon EC2

During the live session, Haowen demonstrated the workflow using an actual Amazon Web Services developer article:

[https://builder.aws.com/content/3CQFzc91nM9sGkfHRbpjo4eXojf/deploying-openclaw-on-amazon-ec2-a-developers-perspective](https://builder.aws.com/content/3CQFzc91nM9sGkfHRbpjo4eXojf/deploying-openclaw-on-amazon-ec2-a-developers-perspective)

In approximately three minutes, Kiro constructed ppt-openclaw-aws-ec2/ containing 17 modular slides categorized across four chapters:

Chapter Slide Contents & Architecture Focus
Part 1: What is OpenClaw Core feature set, business motivation, and deployment value proposition.
Part 2: Architecture Mermaid cloud architecture diagram, infrastructure trade-offs, cost estimates, and IAM prerequisites.
Part 3: Deployment Guide One-click US-region deployment scripts, parameter settings, AWS Systems Manager (SSM) tunnel setup, and multi-region deployment options.
Part 4: Configuration & FAQ Two-column code snippets, operational bash commands, common troubleshooting, and takeaways.

The FAQ slides automatically rendered complex AWS Systems Manager (SSM) commands and bash syntax into neat, dual-column layouts with perfect alignment and dark-mode code syntax highlighting.

+-----------------------------------------------------------------------------------+
EXAMPLE 17-SLIDE DECK STRUCTURE

Cover Page
  --> Part 1: What is OpenClaw     Features and motivation
  --> Part 2: Architecture         Mermaid diagrams and SSM tunnel
  --> Part 3: Deployment Guide     AWS EC2 and regional steps
  --> Part 4: FAQ and Operations   Two-column bash snippets
Closing Page
+-----------------------------------------------------------------------------------+

Deep Dive: Inside .kiro/steering/skill.md

The core intelligence driving this system is .kiro/steering/skill.md. Rather than relying on unpredictable one-shot prompts, Kiro automatically loads this Steering file at the start of every session to enforce strict standards across seven primary domains:

1. Working-Directory Strategy

● Every presentation deck is housed in ppt-{normalized-topic-name}/.

● Reference templates (ppt-aws-theme-demo/ and ppt-aws-theme-demo-cn/) are marked read-only to prevent cross-conversation corruption.

● Theme paths default to theme: ../theme-aws-dark.

2. Intelligent Blog Fetching & Fallbacks

● Executes webFetch(url, mode="full") as primary attempt.

● If returned payload is <500 bytes, indicates a paywall/login gate, or contains only navigation headers, it automatically upgrades to mode="rendered" (running a full browser engine capable of executing JavaScript).

● Hard Rule: Never fabricate or hallucinate content if fetching fails. Inform the user directly and request text input.

3. Language & Internationalization (i18n) Rules

● Automatically sets lang: zh for predominantly Chinese text; otherwise defaults to lang: en.

● When lang: zh is active, the theme switches body font stacks to PingFang SC, HarmonyOS Sans SC, and Noto Sans SC, while applying local font overrides to prevent layout breaks on missing Google Web Fonts.

4. Trademark Compliance Matrix

To maintain corporate and brand standards in Chinese presentation decks, Kiro applies a replacement lookup table:

Source Term Compliant Chinese Translation Standard
AWS Lambda Lambda
AWS web app Amazon Web Services web application
AWS services Amazon Web Services services
built on AWS built on Amazon Web Services

Full service trademarks such as Amazon Bedrock, Amazon EC2, and Amazon S3 are preserved as-is. Under lang: zh, logos and footers are suppressed automatically by the theme.

5. Eight-Step Execution Protocol

Kiro strictly executes these actions in sequence:

1. Read reference templates (slides.md and pages/*.md).

2. Initialize project directory.

3. Draft full slide content.

4. Split long slide sets (>10 slides) into modular chapter pages.

5. Include static assets.

6. Output local preview CLI commands.

7. Apply natural-language feedback iterations.

8. Finalize for export.

6. Content Density & Layout Standards

● Density Cap: Maximum of 7 bullet points per page. Excess content must split into multi-column or multi-page formats.

● Mermaid Architecture Standards: Diagrams must stick to 3–5 core nodes, keep node labels brief, and apply {scale: 0.7} or {scale: 0.8} scaling parameters to prevent visual overflow.

● Icon Selection: Prioritize standard unicode emojis (🚀, ✅, 💡) over unverified icon packages.

● Image Scaling: Enforce explicit width constraints like {width=500px} or {width=60%} on all embedded assets.

● The Zero-Animation Constraint: Direct ban on Vue animations (v-click, v-clicks, v-motion). Static stability ensures flawless PDF and PPTX file conversions.

7. Automated 20+ Item Quality Checklist

Before handing control back to the user, the agent silently validates its output against a 20+ point checklist (verifying asset widths, layout densities, zero-animation rules, and syntax integrity).


Core Engineering Decisions That Guarantee Stability

Having reviewed cloud development workflows for decades, I found three architectural choices in Haowen’s system particularly brilliant:

1. Steering Files vs. System Prompts

Instead of throwing massive prompts into a chat bar, conventions are written once inside .kiro/steering/skill.md. Just as Shiro Seike demonstrated encoding AWS cloud architecture standards into CLAUDE.md earlier in the Summit, Haowen showed that presentation engineering benefits from identical steering discipline.

2. A Constrained Design System

theme-aws-dark/ intentionally provides only eight layout types and three custom Vue components. By limiting visual choices, the AI agent never gets confused between near-identical layout choices, producing clean, predictable results every time.

3. Preset Architectural Snippets

Rather than asking generative AI to construct complex architecture diagrams from scratch, theme-aws-dark/snippets/architectures/ provides battle-tested Mermaid templates for standard AWS infrastructure (Web Apps, Serverless, AI Agents, Data Pipelines). The agent simply renames labels and adjusts connection nodes.


Expanding the Pipeline: Enterprise & Investor Use Cases

While designed around technical blogs, developers at the Hong Kong Summit Developer Lounge quickly discovered that this workflow processes other structured input formats:

● Meeting Transcripts & Recaps: Convert executive transcripts into clean stakeholder summary decks.

● Startup READMEs & Pitch Decks: Transform open-source GitHub repositories into polished investor updates for private VC meetups.

● Internal Architecture Specs: Turn technical markdown documentations into team engineering briefings.

Because the core engine relies on Markdown, Vue 3, and Vite, every deck remains version-controllable with standard git workflows—no proprietary file locks, no software subscription paywalls, and no visual clutter.


Try It Yourself: Quick Start Guide

To test the workflow on your local workstation, run the following commands:

# Clone the repository
git clone https://github.com/hanyun2019/kiro-slidev-ppt-maker.git

# Navigate into the project directory
cd kiro-slidev-ppt-maker

# Open Kiro IDE and send your prompt:
# "Turn this blog post into a PPT: <YOUR-URL-HERE>"


Final Field Notes from the 2026 Developer Lounge

As the neon lights of the Hong Kong skyline flickered across the waters of Wan Chai, the final takeaway from AWS Hong Kong Summit 2026 became clear: The future of developer productivity is spec-driven and deterministic.

Whether you are building enterprise AI agents on Amazon Bedrock or formatting slides for your local user group, encoding your standards into Steering files turns hours of unpaid labor into simple, reliable three-minute loops.

Reported live from the AWS Hong Kong Summit 2026 Developer Lounge. That concludes our 6-part recap series!