Event 1

Report on “Vietnam Cloud Day 2025: Ho Chi Minh City Connect Edition for Builders (Track 1: GenAI & Data)”

Purpose of the Event

  • Learn about security in GenAI and AI Agents to strengthen enterprise safety.
  • Explore the AI-Driven Development Lifecycle (AI-DLC) and how it applies to software development.
  • Understand how to build a unified data foundation optimized for analytics and AI.
  • Stay updated on the latest GenAI strategies and trends on AWS.

Speakers

  • Jun Kai Loke – AI/ML Specialist SA, AWS
  • Kien Nguyen – Solutions Architect, AWS
  • Tamelly Lim – Storage Specialist SA, AWS
  • Binh Tran – Senior Solutions Architect, AWS
  • Taiki Dang – Solutions Architect, AWS
  • Michael Armentano – Principal WW GTM Specialist, AWS

Key Highlights

Main Content

  1. Unified Data Platform on AWS for AI & Analytics

    • Building an end-to-end data pipeline: ingestion → storage → processing → access → governance.
    • Breaking down silos in data, people, and processes; enabling self-service & standardized governance.
    • Key services: S3, Glue, Redshift, Lake Formation, OpenSearch, Kinesis/MSK.

  2. GenAI Strategy on AWS

    • Vision, trends, and enterprise adoption roadmap.
    • Amazon Bedrock: model selection, RAG, guardrails, cost/latency optimization.
    • AgentCore & Amazon Nova with support for frameworks (CrewAI, LangGraph, LlamaIndex…).

  3. Securing GenAI Applications

    • OWASP LLM risks; multilayered security: infrastructure → model → application.
    • Five pillars: Compliance, Privacy, Controls, Risk Management, Resilience.
    • Tools: Bedrock Guardrails, Human-in-the-loop, Observability (OpenTelemetry).

  4. AI Agents – Productivity Boosters

    • From assistants to multi-agent systems, automation with less supervision.
    • Use cases: customer support, BI with Amazon Q (QuickSight), process automation.

  5. Reliability & Accuracy of GenAI

    • Mitigating hallucination with Prompt Engineering, RAG, Fine-tuning.
    • RAG workflow: input → embedding → context → LLM → output.

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

    • Lifecycle: Inception → Construction → Operation.
    • Evolution: AI-Assisted → AI-Driven → AI-Managed.
    • Implementation with IaC, automated testing, monitoring, and risk management.

  7. Amazon SageMaker – Unified Studio

    • Unified environment for data, analytics, and AI.
    • Supports Lakehouse, governance, Zero-ETL integration (S3 ↔ Redshift, Aurora, DynamoDB, RDS…).
    • Full MLOps: pipelines, registry, deployment, monitoring.
    • Integrated with Bedrock & JumpStart to accelerate GenAI application development.

Key Learnings

  • Design Mindset

    • Build data & AI systems end-to-end, removing silos.
    • Apply self-service and governance principles from the start.

  • Technical Architecture

    • Integrate AWS services (S3, Glue, Redshift, SageMaker, Bedrock…) into a unified platform.
    • Apply Zero-ETL, Lakehouse, MLOps for scalability, governance, and sustainable operations.
    • Leverage AI Agents and GenAI frameworks to automate processes and boost productivity.

  • Strategy

    • Define a GenAI adoption roadmap balancing innovation speed and cost.
    • Focus on multilayered security: infra, model, application; combine guardrails & human-in-the-loop.
    • Prioritize reliability and accuracy with RAG, prompt engineering, fine-tuning.

  • Software Development Mindset

    • Transition from AI-Assisted → AI-Driven → AI-Managed.
    • Adopt AI-DLC to standardize development with AI involved at every stage.

Application to Work

  • In projects:

    • Experiment with AI Agents for registration/login and customer support.
    • Use validation/guardrails to safely integrate GenAI into applications.

  • In learning & team projects:

    • Apply AI-DLC for task division: AI supports code/docs generation, team reviews & approves.
    • Know when to use Lambda (serverless) vs containers (ECS/Fargate).

  • As an intern:

    • Learn to apply a business-first approach when writing documentation or gathering requirements.
    • Realize the importance of a solid data foundation for GenAI to deliver real value.

Event Experience

Joining the “GenAI-powered App-DB Modernization” workshop was a highly valuable experience, giving me a holistic view of modernizing applications and databases using cutting-edge methods and tools. Some key takeaways:

Learning from Experts

  • AWS experts shared the latest trends in GenAI, Data Foundation, and Security.
  • Gained a clearer understanding of building a unified data foundation for AI & Analytics.
  • Impressed by the vision of AI Agents and their potential to enhance productivity.

Hands-on Technical Insights

  • Learned how to design an end-to-end data pipeline: ingestion → storage → processing → access → governance.
  • Explored tools like Amazon Bedrock, AgentCore, and SageMaker Unified Studio.
  • Discovered solutions to reduce hallucination (Prompt Engineering, RAG).
  • Understood how to apply AI-DLC for balancing tasks between AI and humans in software development.

Tools & Methods in Practice

  • Explored Bedrock Guardrails to ensure safe GenAI implementation.
  • Understood when to use serverless (AWS Lambda) vs containerization (ECS/Fargate).
  • Learned how to leverage Amazon Q for BI (QuickSight) and customer support.

Networking & Exchange

  • The event was a great chance to interact with AWS experts and learn from real-world case studies.
  • Realized the importance of a business-first approach in every technology decision.

Key Takeaways

  • GenAI is not just a tool, but requires the right strategy and architecture to generate value.
  • Data and security are the foundations—without them, AI cannot thrive.
  • AI Agents and AI-DLC are set to reshape how we design and operate systems.

Event Photos

(Add photos from the event here)