Event 2

Workshop Report: “Platform Engineering & GenAIOps”

Event Objectives

  • Explore the role of Platform Engineering in the modern Cloud ecosystem and its implications for career development.
  • Gain a foundational understanding of GenAIOps principles and DevOps practices specific to Generative AI applications.
  • Learn about emerging tools and strategies for efficient code deployment in the Agentic Era.
  • Understand the architectural considerations for building production-ready Multimodal GenAI solutions on AWS.

Speakers

  • Hai Bui - Engineering Manager, GoTymeX
  • Phuc Dang - Cloud Architect, GoTymeX
  • Phap Nguyen - Cloud Engineer, VPBank
  • Phat Pham - Software Engineer, Katalon
  • Nghi Danh - AI Engineer, Renova Cloud
  • Phong Nguyen - Senior Software Engineer, Sympli
  • Thinh Nguyen - DevOps Engineer, FCAJ

Key Highlights

Building Modern Platform Engineering and Career Orientation

  • Introduction to Platform Engineering and its indispensable role within the broader DevOps ecosystem.
  • Insights into organizational culture, internship pathways, and direct Q&A sessions with working professionals in the field.

GenAIOps Essentials - DevOps for Generative AI Applications

  • Foundational principles of DevOps on AWS and relevant learning resources for practitioners.
  • Practical demonstrations of GenAIOps applied to AWS projects, including the use of Bedrock AgentCore Observability, Amazon EKS, and Langfuse for monitoring and evaluation.

Code Deployment in the Agentic Era

  • Analysis of current challenges in software development workflows and the introduction of AI-assisted tooling to address them.
  • An in-depth exploration of the Productivity Playbook with live demonstrations of AI-assisted code authoring.

Production-Ready Multimodal GenAI on AWS

  • Introduction to the modern AI application stack and its key architectural components.
  • Multimodal Search powered by Nova Embeddings for enhanced data retrieval across diverse content types.
  • Leveraging GraphRAG for enterprise knowledge management and designing scalable Multi-Agent Workflows.
  • Strategies for ensuring safety, observability, and reliability in GenAI systems operating in production environments.

Key Takeaways

Technical Thinking

  • Platform engineering serves as the critical foundation for scalable operations in the Cloud. The transition to the Agentic Era demands a fundamental rethinking of how code is authored, reviewed, and deployed, with AI-assisted tooling playing an increasingly central role in productivity.

System Architecture

  • Transitioning AI from experimental environments to production requires a disciplined GenAIOps approach. Combining Nova Embeddings, GraphRAG, and Multi-Agent workflows provides a flexible and powerful architecture capable of handling complex, real-world business data at scale.

Applying to Work

SmartInvoice Shield Project:

  • Integrating Multimodal Search with Nova Embeddings: Enhance the system’s capability to extract and cross-verify data from complex invoice formats, including blurry photographs and text-heavy PDF documents.
  • Implementing Multi-Agent Workflows: Apply an AI Agent architecture — analogous to the OpenClaw model — to independently handle discrete steps in the invoice processing pipeline, such as a dedicated OCR reading agent and a specialized cross-verification and fraud detection agent.
  • Applying GenAIOps: Utilize Bedrock AgentCore Observability and Langfuse to monitor the performance and accuracy of underlying AI models, ensuring stable and reliable system operation.

Event Experience

Learning from Multi-disciplinary Experts

  • Listening to practitioners from GoTymeX, VPBank, and Katalon provided a multi-dimensional perspective on how Cloud and AI technologies are applied across different industries and organizational contexts. The diversity of backgrounds among speakers significantly enriched the breadth of knowledge shared throughout the event.

Accessing Practical Architecture Patterns

  • Demonstration sessions on Multimodal GenAI architecture helped bridge the conceptual gap between basic API integration and the design of a genuinely enterprise-grade AI system capable of handling production workloads.

Expanding Professional Connections

  • Open Q&A sessions and designated networking breaks offered substantive opportunities to engage with speakers and peers, gain insights from real-world deployment experience, and receive actionable advice relevant to ongoing work in Cloud system development.

In summary, this workshop provided a comprehensive view of where the industry is heading in terms of AI-augmented development and cloud-native architecture. The knowledge and connections gained have directly informed how I approach both technical problem-solving and professional development during my internship.