Event 2

AI-DRIVEN DEVELOPMENT WORKSHOP — Shaping the Future of Development**

Location: AWS Event Hall, 26th Floor — Bitexco Tower, Ho Chi Minh City

Time: 14:00 – 16:30, Friday, 3 October 2025

Speakers: Toan Huynh, My Nguyen

Organizers: Diem My, Dai Truong, Dinh Nguyen

Event objectives

  • Share new trends in AI-driven software development (AI-Driven Development).
  • Present the AI-Driven Development Lifecycle (AI-DLC) — a model integrating AI across the software lifecycle.
  • Demo two important tools: Amazon Q Developer and Kiro IDE extension.
  • Analyze how AI improves productivity, speed and product quality.

AI-DLC: three evolution stages

  • AI-Assisted Development: AI supports code generation, suggestions and syntax checks.
  • AI-Driven Development: AI participates in architecture, planning and decision support.
  • AI-Managed Development: AI orchestrates parts of the delivery pipeline with human approval.

In this model AI acts as a smart coordinator, while humans retain final decision and validation responsibilities.

Benefits of AI in software development

  • Predictability: Maintain schedules and better forecast release timelines.
  • Velocity: Accelerate time-to-market for ideas.
  • Quality: Reduce defects and increase stability.
  • Innovation: Enable idea exploration and new approaches.
  • Developer engagement: Improve developer satisfaction and effectiveness.
  • Customer satisfaction: Enhance user experience and trust.
  • Productivity: Reduce development time and increase overall throughput.

Standard AI‑DLC workflow

The workshop presented a four-step workflow:

  • Requirement: Product Owner gathers and analyzes requirements.
  • Design: Software Architect defines system design and APIs.
  • Implementation: Software Engineers develop, test and integrate.
  • Deployment: Release and monitor the system.

AI supports all phases to make them more consistent, traceable and efficient.

Key workflow features

  • Role separation: Clear separation between product, architecture and implementation.
  • AI‑Enhanced: Each role has AI personas tailored to the responsibility.
  • Iterative: Continuous feedback loops between stages.
  • Template‑driven: Standardized outputs using AI‑DLC templates.

AI in each development stage

  • Specific context: Roles (PM, Architect, Developer) get role‑specific AI assistants.
  • Clear inputs/outputs: Define precise inputs and expected outputs between steps.
  • Interactive approach: Human-AI collaboration with real-time feedback.
  • Documentation: Keep prompts.md and dashboard.md updated to track progress and improvements.

Demos: Amazon Q Developer & Kiro IDE

Amazon Q Developer

  • AI assistant integrated into IDEs (VS Code, Cloud9, …).
  • Automates code generation, testing, documentation and AWS architecture suggestions.
  • Helps update prompt.md and automates parts of CI/CD.
  • Demo showed planning, generating user stories and managing project tasks via AI.

Kiro IDE (demo by My Nguyen)

  • Extension for generating and managing spec documents (requirements.md, design.md, tasks.md).
  • AI can scaffold feature descriptions, API flows and backend code.
  • Demo included building a Chat app with AI-driven auth flows (signup, login, token handling).

Lessons learned

  • AI complements developers as a “smart teammate”, not a replacement.
  • AI-DLC provides consistency and transparency across the lifecycle.
  • Amazon Q Developer saves time, reduces defects, and helps automate testing and deployment.
  • Kiro IDE demonstrates end-to-end support from spec to code.
  • Integrating AI into DevSecOps is essential for high productivity and secure delivery.

Practical applications

  • Use Amazon Q Developer to automate internal test generation and documentation.
  • Use Kiro IDE to standardize spec workflows and speed backend development.
  • Run AI‑driven sprints to evaluate AI impact on team delivery.
  • Adopt “AI-Assisted Code Review” to improve product quality.

Personal impressions

The “AI-Driven Development Workshop” was a focused, practical session that clarified how AI can be embedded across the software lifecycle. Presentations by Toan Huynh and My Nguyen provided actionable guidance and real demos that are directly applicable to our projects.