John Ritsema
Principal Solutions Architect, AWS
Actions
John Ritsema is a Principal Solutions Architect at AWS. He enjoys working with developers to help them build highly available, scalable, and resilient applications. He is currently focused on AI, containers, serverless, and cloud native technology.
Agentic AI with Model Context Protocol (MCP): Building Smarter AI Assistants
Large Language Models are incredibly powerful, but they're fundamentally limited by their training data cutoff dates and lack of access to real-time information. While RAG (Retrieval-Augmented Generation) helps inject static context, modern AI assistants need something more dynamic - the ability to interact with live systems, APIs, and data sources in real-time.
Enter the Model Context Protocol (MCP), an open standard that's revolutionizing how AI agents connect to external tools and data sources. MCP provides a unified protocol that eliminates tight coupling and enables seamless interoperability.
**What You'll Learn:**
- **MCP Architecture Deep Dive**: Explore the building blocks - MCP hosts, clients, servers, and transport mechanisms (stdio and streamable http)
- **Protocol Capabilities**: Learn how MCP servers expose tools, resources, and reusable prompts to AI agents
- **Server Development**: Walk through an example MCP server for AWS Lambda operations using official SDKs, including tools for listing functions, updating runtime versions, and safe deployment practices
- **AWS Native Solutions**: Discover Amazon Bedrock AgentCore services and how they leverage MCP for scalable agent deployments
Walk away with practical knowledge and code samples to start building your own MCP servers and creating more capable, context-aware AI assistants.
Building developer platforms with AWS Proton
Come learn about automated CI/CD and IaC pipelines and how to scale them in your organization using self-service templates.
From Chat to Agent: Building Scalable AI with Amazon Bedrock AgentCore
Traditional AI chatbots face challenges when scaling: session management complexity, memory persistence overhead, and limited observability. Amazon Bedrock AgentCore changes this paradigm with purpose-built infrastructure for AI agents.
In this session, you'll learn how to transform a traditional RAG chatbot into a robust AI agent using AgentCore's key services. We'll explore:
• Agentic RAG - Moving from static retrieval to dynamic tool-calling patterns
• Session isolation - How dedicated containers per user session simplify state management and enhance security
• Managed memory - Automatic conversion of short-term conversations into long-term insights and user preferences
• Serverless vector storage - Eliminating database overhead with S3 Vectors
• Built-in observability - End-to-end tracing from user interactions to LLM calls
Through a demo and walkthrough of the AI Agent Accelerator reference implementation written in Python and the Strands Agent framework, you'll see how AgentCore Runtime enables 8-hour agent sessions, how AgentCore Memory eliminates database complexity, and how S3 Vectors provides truly serverless vector storage.
Walk away with a GitHub repository containing the complete reference architecture and deployment scripts to deploy scalable AI agents on AWS.
Making web development fun again with HTMX
Have you grown weary of the complexity of building web apps today? In this talk, I'll introduce you to a JavaScript library called HTMX that, ironically, allows you to build dynamic web apps without writing JavaScript. HTMX combined with Tailwind CSS and Go makes for a refreshingly simple, fun, and maintainable web development stack. I'll walk through a sample application and show how you can deploy it to AWS.
Previewing PR environments using containerized AWS Lambda functions
In this talk, I'll show how to leverage containers on AWS Lambda Function URLs for extremely fast, low-cost, ephemeral, web app preview environments that can be provisioned in seconds for every pull request and commit. These preview environments scale-to-zero and you only pay for actual HTTP requests.
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top