AI-301
Agentic AI Cloud – Scalable Stateful Infrastructure - Mastering Cloud Native MCP Server & API Development and Deployment
Master the deployment of scalable, stateful AI agents using Kubernetes, Docker, and Dapr. Learn to build and observe Cloud Native MCP Servers and APIs with robust architecture, context storage, and design thinking for production-grade Agentic AI systems.
Available Sections:
Details
AI-301 represents the pinnacle of the Agentic AI Engineering series, uniquely focusing on the deployment of stateful and scalable AI Agents using Docker, Kubernetes, Dapr, and Cloud Native Model Context Protocol (MCP) Servers and APIs. (The upcoming version of MCP servers will support remote cloud deployment in addition to the current on‑premise setup.)\nThis intensive course equips students with the specialized skills to design, build, deploy, and scale highly performant, robust AI Agents in the cloud and cloud-native MCP infrastructure essential for advanced Agentic AI systems. You will master the complete lifecycle of creating production-ready, cloud-native MCP Servers and APIs, from backend development to cloud deployment, user-centered design, and robust operational practices. Learn to leverage a cutting-edge technology stack specifically to build scalable and efficient Cloud Native AI Agents and Cloud Native MCP solutions that underpin the next generation of intelligent agent applications.
Key Learning Modules
Module 1Dapr, Dapr Workflows, and Dapr Agent
Provide the infrastructure for deploying and scaling long-term, stateful agentic workflows. This managed platforms abstract execution, state persistence, and tool integration, with tracing. Eliminates the need for bespoke cloud setups, offering scalability and reliability for multi-agent systems running over weeks or months. In this module we will become masters of Agentic AI deployments and observability. For short-term, transient workflows, learn to use the OpenAI Agents SDK’s handoff mechanism, while for long-term, stateful workflows, the Dapr approach — complemented by OpenAI Agents SDK — offers the best solution.
Module 2Mastery of Kubernetes and Dapr
Provides the infrastructure for deploying and scaling long-term, stateful agentic workflows in the Cloud. This managed platforms abstract execution, state persistence, and tool integration, with tracing.
Module 3Foundation of Cloud Native MCP Servers
This module establishes the foundation for building Cloud Native MCP Servers using FastAPI, SQL, and GQL. Students will learn to design high-performance API endpoints specifically for MCP server functionalities, focusing on efficiency, scalability, and cloud-native best practices. Explore SQL and GQL to enable flexible and efficient data retrieval and manipulation within MCP APIs, crucial for agent context management and data exchange.
Module 4Building Scalable MCP Context Storage with Graph Databases
Master Neo4j to build robust and scalable context storage for Cloud Native MCP Servers. Learn to design graph-based data models optimized for storing and retrieving agent context, knowledge graphs, and long-term memory within an MCP architecture. Focus will be on leveraging Neo4j within a cloud-native MCP server to ensure performance and scalability.
Module 5Integrating Agent to Agent communication using Agent to Agent Protocols (AI-to-AI)
Master the integration of AI Agents and message passing.
Module 6Design Thinking for AI-Centric Agent Experiences with MCP
Integrate Design Thinking methodologies to ensure Cloud Native MCP Servers and APIs effectively support AI-centric agent experiences. Learn to design MCP and API functionalities that directly address user needs and AI Agent requirements, focusing on optimizing the MCP's role in enabling valuable and AI-friendly Agentic AI interactions within a cloud-native context.
Module 7Behavior-Driven Development (BDD) for Robust Cloud MCP Servers
Implement Behavior-Driven Development (BDD) to ensure the robustness, reliability, and predictable behavior of Cloud Native MCP Servers. Learn to define MCP API behavior in AI-centric language, write executable specifications, and automate testing to guarantee MCP servers function correctly and meet the performance and reliability demands of cloud based Agentic AI applications.
Course Outcomes
Deploy and observe stateful longrunning AI Agents on Kubernetes.
Design and develop high-performance, scalable Cloud Native MCP APIs using FastAPI, SQL and GQL.
Build and integrate scalable context storage for Cloud Native MCP Servers using Postgres and Neo4j.
Containerize and deploy Cloud Native MCP Servers locally and to the cloud using Docker, and Kubernetes.
Apply Design Thinking to create AI-centric Cloud Native MCP and API Server functionalities for enhanced agent experiences.
Utilize Behavior-Driven Development (BDD) to ensure the robustness and reliability of Cloud Native MCP APIs in production environments.
Prerequisites
Note: These prerequisites provide essential knowledge for success in this course. If you haven't completed these courses, consider taking them first or reviewing the relevant materials.