Event 2

Learning Notes from “AI-Driven Development Lifecycle: Reinventing Software Engineering”

Event Objectives

  • Understand how AI can automate and optimize each stage of the Software Development Lifecycle (SDLC).
  • Grasp the philosophy of AI augmenting humans rather than replacing them in the application development process.
  • Directly observe how Amazon Q and other AI tools support developers from ideation, coding, to infrastructure deployment (Infrastructure as Code – IaC).
  • Recognize the “AI-first development” trend – where AI becomes a natural part of future software development processes.

Speaker List

  • Mr.Toan Huynh - PMP, Senior Solutions Architect, AWS
  • Ms.My Nguyen - Senior Solutions Architect, AWS

Highlighted Content

Challenges in AI-Assisted Programming

The opening section presented limitations and challenges when integrating AI into programming:

  • AI cannot yet handle projects with complex logic requiring deep understanding of business context.
  • Developers struggle to control details in generated code if they don’t clearly describe objectives and scope.
  • Code quality heavily depends on prompts and context provided by users.

This is precisely why AI-DLC was created: to establish a structured process enabling more effective collaboration between AI and humans.

AI in Development – How AI is Changing Software

This section analyzed how AI is transforming the software industry:

  • AI assists in code generation, technical documentation, API design, and automated testing.
  • Developers transition from “code writers” to “AI orchestrators” — coordinators who guide and evaluate outputs.
  • Tools like Amazon Q, GitHub Copilot, ChatGPT for Developers become central tools in modern dev team workflows.

Introduction to AI-DLC

AI-Driven Development Lifecycle (AI-DLC) is a software development approach with AI collaboration, where each step is designed to provide AI with specific context and objectives to generate more accurate results.

Inception

  1. Build Context on Existing Codes – AI is “fed” with current source code to understand project structure.
  2. Elaborate Intent with User Stories – Developers describe requirements through user stories, clarifying objectives.
  3. Plan with Units of Work – Break down work into small units for AI to execute and generate code incrementally.

Construction

  1. Domain Model (Component Model) – Build domain models or logical architecture diagrams.
  2. Generate Code & Test – AI generates code and automated tests based on planned information.
  3. Add Architectural Components – Add architectural components like APIs, data layers, logging, security.
  4. Deploy with IaC & Tests – Automatically deploy systems with Infrastructure as Code and integration tests.

Each step provides additional “rich context” for the next step, helping AI understand the system more deeply and generate increasingly accurate results.

CORE CONCEPTS – Three Core Principles

  1. Context Awareness – AI needs clear context about code, requirements, and domain to operate effectively.
  2. Collaborative Generation – Humans and AI collaborate: AI generates code, humans guide and review.
  3. Continuous Refinement – Iterative process to refine outputs and improve quality.

Mob Elaboration

Mob Elaboration is a method for expanding requirements (intent elaboration) through team collaboration:

  • Multiple members jointly describe requirements, ask questions, and add information for AI.
  • Helps AI understand more deeply about business, objectives, and complex project logic.
  • This approach helps reduce misunderstanding risks, especially in large or multi-domain teams.

5-Stage Sequential Process of AI-DLC

AI-DLC is executed through 5 stages:

  1. Inception – Understand requirements, analyze systems.
  2. Construction – Create domain models and initial structure.
  3. Generation – Automatic code generation.
  4. Testing – Automate unit and integration testing.
  5. Deployment – Deploy applications with IaC and CI/CD pipelines.

Each iteration helps AI learn more and improve output quality.

Demo 1 – Hands-on Experience with AI DLC using Amazon Q

The demo illustrated how to apply AI-DLC in practice through a small project:

  • Starting from a simple idea → converting to user story describing business requirements.
  • AI assists in dividing work (Units of Work) and detailed planning for each module.
  • Participants can control AI through questions, checkboxes, and logical conditions, helping AI understand the scope of work.
  • AI continues to generate code, write tests, create project structure, and automatically deploy trials.
  • The demo clearly showed how AI and humans collaborate harmoniously: AI handles repetitive tasks, humans guide and make strategic decisions.

Introduction to Kiro

Kiro’s Philosophy

The next part of the workshop introduced Kiro, an intelligent development environment designed around the “AI-native development” philosophy – where AI is a core component, not just a support tool.

Kiro’s philosophy focuses on three main elements:

  1. Deep integration with development processes – AI not only assists in writing code but also participates in planning, managing context, and analyzing change impacts.
  2. Comprehensive project context understanding – Kiro maintains continuous state awareness of system structure, allowing AI to interact with the entire project rather than individual files.
  3. Intelligent control & collaboration – Developers can guide AI through contextual commands, ensuring each change has a clear purpose and is consistent with the system.

Project Structure in Kiro

Unlike traditional text editors like VSCode or JetBrains, Kiro is not just a code writing environment — it’s an AI workspace with structural awareness.

Project structure in Kiro includes:

  • Context Layer – Stores context, domain models, and relationships between modules.
  • Task Layer – Manages Units of Work tracked and gradually completed by AI.
  • AI Agent Layer – Each task (code, test, refactor, deploy) has a dedicated agent, creating a multi-agent – collaborative – parallel development model.
  • Human-in-the-Loop Control – Developers can intervene at every step: confirm, modify, or reject AI outputs.

This makes Kiro not just a code generation tool but a collaborative development ecosystem between humans and AI.

Demo 2: Kiro – Applying AI-DLC

In the demonstration, the speaker illustrated how Kiro operates AI-DLC seamlessly:

  1. User inputs a basic business requirement, e.g., “build an event management system”.
  2. Kiro automatically analyzes intent, creates domain models, and breaks down into user stories.
  3. AI in Kiro generates corresponding modules, components, and test cases.
  4. Developers can interact through checkbox-based task control to confirm each part of the work.
  5. Finally, Kiro deploys the complete system with IaC and automated testing.

The demo showed that AI-DLC is not just theory, but can be implemented practically within the Kiro environment — where AI, humans, and development processes merge into a unified system.

Event Experience

Participating in the “AI DLC x Kiro: Reinventing Developer Experience with AI” workshop was an extremely valuable experience, helping me better understand how AI is deeply integrated into software development environments and how Kiro’s design philosophy brings a new approach to developers.

Learning from Expert Speakers

  • Speakers shared about AI DLC – a platform supporting AI-based software development, automating many SDLC processes.
  • Additionally, the introduction to Kiro Editor provided deep insights into building a text editor in an AI-native direction rather than just “adding AI plugins” to old environments.
  • I was particularly impressed with Kiro’s philosophy: minimalist, high-performance, focused on user experience and module-based scalability.

Practical Technical Experience

In learning:

  • Apply AI-DLC structure for personal projects
  • Practice “Context Awareness” principle with AI assistants
  • Build habit of writing clear requirements as user stories

For future career:

  • Understand modular, extensible, maintainable system design like Kiro
  • Master Amazon Q and other AI tools effectively
  • Recognize importance of providing quality context for AI

Mindset shift:

  • Approach problems with “AI-augmented” thinking
  • Consider building custom tools with deep AI integration
  • Always ask: “How can AI assist better at this step?”

Applying Modern Tools

  • Experiencing AI DLC on Kiro helped me better understand the capability of automating development processes, especially in steps like code generation, documentation, and debugging.
  • I recognized the potential of building personal learning and working tools with intelligent suggestions, helping shorten development time and improve product quality.
  • Kiro’s modular design concepts also suggested directions for designing flexible, scalable, and maintainable systems.

Networking and Exchange

  • The workshop created opportunities for me to connect with developers, AI researchers, and product designers, thereby learning more about the AI-augmented development trend.
  • Through discussions, I learned much about how AI can play the role of creative collaborator, helping developers focus more on logic and system thinking rather than repetitive operations.

Lessons Learned

Participating in the “AI DLC x Kiro” workshop was a turning point in how I perceive AI’s role in software development.

The most important thing I learned wasn’t specific tools, but the necessary mindset shift:

  • AI is not a tool for faster coding
  • AI is a partner for better thinking and system design
  • Structured processes (like AI-DLC) are more important than raw AI power

The workshop also showed me the future of development tools - where AI-first architecture like Kiro will become standard, and developers need to prepare for this paradigm shift.

Insights from AWS Solution Architects and hands-on experience with Kiro equipped me with a solid foundation to apply AI in my learning journey and future career in software engineering.

Some Photos from the Event

Event_02

Group photo check-in after the event

This is the group check-in moment after the workshop ended. This event provided many valuable insights about how AI is reshaping development workflow.

Event_02

Professional event space

The workshop was professionally organized with complete demo stations and networking opportunities. This was one of the important events that helped me deeply understand AI-driven development.