Event 2

Summary Report: “AI-Driven Development Life Cycle: Reimagining Software Engineering”

Event Objectives

Explore how AI is transforming the entire software development lifecycle, from planning to deployment. Learn how to integrate AI to increase productivity and focus on high-value creative tasks.

Event Information

  • Time: 2:00 PM - 4:30 PM, Friday, October 3rd, 2025
  • Location: AWS Event Hall, L26 Bitexco Tower, HCMC
  • Speakers: Toan Huynh, My Nguyen

Key Highlights

1. AI in Development - Outcomes

AI adoption brings:

  • Velocity - Reduce time-to-market
  • Quality - Meet usability & reliability expectations
  • Market Responsiveness - React quickly to market changes
  • Innovation - Drive innovation
  • Developer Engagement - Increase developer satisfaction
  • Productivity - Increase value, reduce costs

2. Challenges with AI Development

AI-Managed: Unreliable, hard to explain, lacks flexibility

AI-Assisted: Not truly effective, manual inefficiencies, technical debt accumulation

3. AI-Driven Development Lifecycle (AI-DLC)

Core Concept:

  • AI as Collaborator: AI assists developers, humans control critical decisions
  • Human-Centric: Developers remain central, AI enhances not replaces
  • Accelerated Delivery: Development cycles reduced from weeks/months to hours/days

Two phases:

  1. Inception: Build Context → User Stories → Plan
  2. Construction: Domain Model → Generate code → Deploy with IaaC

4. 5-Stage Sequential Process

  1. Product Owner → 2. Architect (Design) → 3. Architect (Construction) → 4. Engineer (POC) → 5. Engineer (MVP)

5. Anti-Patterns - 7 Things to Avoid

  1. Don’t single-shot multi-step problems
  2. Maximize semantics-to-token ratio
  3. Refresh context strategically
  4. Control AI over-reach
  5. Model knows old better than new
  6. Brownfield needs special context building
  7. Think surgical precision

6. Amazon Q Developer

Prompt Structure: Role → Plan (markdown with checkboxes) → Task

Example: Build travel booking app with AI integration

Workflow: Create folder → User stories → Clarification → Checkbox → Review → Execute

7. Kiro - AI-Powered Coding Assistant

Kiro is AWS’s AI coding assistant with 4 key features:

  1. Agent Hooks: Auto-trigger tasks on events (file save), generate docs/tests/optimize
  2. VS Code Compatible: Support plugins, themes, settings
  3. Claude Models: Sonnet 3.7/4 with powerful reasoning
  4. Enterprise Security: Built & operated by AWS

Advantages: High automation, context-aware, documentation-driven, enterprise-ready

Key Takeaways

Participating in the “GenAI-powered App-DB Modernization” workshop was an incredibly enriching experience, providing me with a comprehensive perspective on modernizing applications and databases using contemporary methods and tools. Some notable highlights include:

Learning from highly skilled speakers

  • Speakers from AWS and major technology organizations shared best practices in modern application design.
  • Through real-world case studies, I gained a deeper understanding of how to apply Domain-Driven Design (DDD) and Event-Driven Architecture to large-scale projects.

Hands-on technical exposure

  • Participating in event storming sessions helped me visualize how to model business processes into domain events.
  • Learned how to split microservices and define bounded contexts to manage complexity in large systems.
  • Understood the trade-offs between synchronous and asynchronous communication, as well as integration patterns like pub/sub, point-to-point, and streaming.

Leveraging modern tools

  • Directly explored Amazon Q Developer, an AI tool supporting the SDLC from planning to maintenance.
  • Learned how to automate code transformation and pilot serverless with AWS Lambda, thereby enhancing development productivity.

Lessons learned

  • Applying DDD and event-driven patterns helps reduce coupling while increasing scalability and resilience for systems.
  • Modernization strategies require a phased approach and ROI measurement; rushing to transform the entire system should be avoided.
  • AI tools like Amazon Q Developer can boost productivity when integrated into the current development workflow.

Some event photos

event-2