Event 1

Workshop Report: “AWS re:Invent Recap HCMC 2026”

Event Objectives

  • Tech Updates: Capture the latest announcements from global AWS re:Invent, with a specific focus on the Nova model series, Amazon Bedrock, and MLOps infrastructure.
  • Solution Optimization: Explore practical implementations of Vector Databases and Multimodal Retrieval to address complex data challenges.
  • Local Engagement: Facilitate direct networking with AWS experts on topics of application modernization and Vietnam’s new DirectConnect POP infrastructure.

Speakers

  • Prapti Gupta - AWS Speaker and Expert (Led the session on Vector Database on S3)
  • Solutions Architects from AWS Vietnam

Key Highlights

Modern Data Platform and Analytics

  • Keynotes on ML/AI and Innovation: AWS’s strategic vision for driving large-scale innovation through the convergence of data and artificial intelligence.
  • Vector Database on Amazon S3: A method for storing and managing data vectors directly on S3, designed to optimize cost and scalability for AI-driven applications.
  • Natural Query Language in OpenSearch: Enabling users to query complex data systems using natural language, significantly reducing the technical barrier between end users and data infrastructure.

Generative AI Ecosystem and MLOps Infrastructure

  • Upgrading GenAI Applications: Leveraging the latest capabilities of the Nova model family and Amazon Bedrock to develop more intelligent and context-aware applications.
  • Resource Management via MCP: A structured mechanism for efficient interaction and coordination of AWS resources through the Model Control Plane.
  • Multimodal Retrieval: A significant advancement in Bedrock Knowledge Bases, enabling search and synthesis of information across multiple data types including images, text, and tabular data.
  • Large-scale MLOps Infrastructure: Professional-grade workflows for training and deploying machine learning models at scale on the Amazon SageMaker AI platform.

Key Takeaways

Professional Knowledge

  • Architecture Flow: Gained a comprehensive understanding of the end-to-end data processing pipeline — from storage on S3 and querying via OpenSearch, through to model training and inference on SageMaker and Bedrock.
  • Power of Multimodal: Understood how combining information from multiple data formats enhances the accuracy and robustness of AI systems.

Systems Thinking

  • Developed a clearer recognition that effective AI technology extends beyond individual models — it encompasses the entire infrastructure ecosystem and governance framework (MLOps) required to ensure operational scalability and reliability.

Applying to Work

SmartInvoice Shield Project:

  • Apply Multimodal Retrieval to handle complex and varied invoice inputs, including blurry photographs and structured PDF documents.
  • Evaluate replacing or supplementing relational databases with Vector Databases to accelerate cross-referencing and enable intelligent, context-aware search.
  • Establish standardized MLOps workflows to manage OCR model versioning and deployment more systematically.

Event Experience

Attending the AWS re:Invent Recap at the Bitexco Financial Tower was both a professionally enriching and inspiring experience. The venue and event format reflected the high caliber of participants and content.

Direct Engagement with Subject Matter Experts

  • Listened to presentations delivered by experienced speakers on optimal cloud storage architectures and the evolving role of AI in modern data platforms.
  • The sessions on Vector Databases and Multimodal Retrieval were particularly relevant to ongoing project work, providing concrete implementation guidance beyond theoretical concepts.

Networking with the Cloud and AI Community

  • The event offered meaningful opportunities to connect with engineers, architects, and practitioners across Vietnam’s Cloud and AI ecosystem.
  • Conversations during breaks and Q&A sessions allowed me to validate technical assumptions and gain perspective from professionals with hands-on production experience.

Lessons Learned

  • Architecture over tools: Understanding the flow of data through a system is more valuable than familiarity with any individual service in isolation.
  • Cost-conscious design: Optimizing for cost from the design phase — such as leveraging Amazon S3 for Vector Database storage — is a critical consideration for the long-term sustainability of large-scale Cloud projects.
  • Professional networking as a learning tool: Engaging with peers and experts is not merely a social activity, but an effective mechanism for stress-testing one’s own technical understanding against real-world experience.

Event Photos

Sneak peek at the event

Sneak peek at the event

Overview of the event

Overview of the event

Overall, this event reinforced the importance of staying current with the rapidly evolving AWS ecosystem and provided a structured framework for applying cutting-edge AI and data platform technologies to real-world engineering challenges.