About This Course
Welcome to the "Al Agent - Fundamental in Building Agents" course, a two-day, hands-on program for IT professionals, enterprise software engineers, and automation consultants who want to move from basic prompt engineering to building robust, interconnected agent architectures. It is especially valuable for technical leads and system integrators implementing both low-code and code-driven Al automation across enterprise, factory, and development environments.
Over two days you will master the full spectrum of agent design - from no-code workflow automation in n8n to framework-driven architectures with Semantic Kernel and LangGraph. You will securely integrate systems using the Model Context Protocol (MCP) and the Agent-to-Agent (A2A) protocol, build controllable stateful workflows, trace and evaluate agents with LangSmith, and complete a capstone where you design, build, test, and present a working agent prototype.
Learning Objectives
Upon completing this course, you will be able to:
- Distinguish between chatbots, RAG, workflow automation, and Al agents, and explain the core components and the agent loop.
- Evaluate business tasks for agentic feasibility, and decide when fixed automation versus agentic decision-making is appropriate.
- Design agent workflows by mapping inputs, outputs, decisions, tool actions, and human-in-the-loop approval points.
- Build agents using no-code/low-code tools such as n8n, understanding both their capabilities and limitations.
- Securely integrate tools and data using the Model Context Protocol (MCP), including building and connecting an MCP server.
- Enable agent interoperability using the Agent-to-Agent (A2A) protocol and connect collaborating agents.
- Build enterprise-style agents with Semantic Kernel using plugins, functions, planners, and
- Design stateful, controllable agent workflows with LangGraph using nodes, edges, state, and conditional routing
- Evaluate, trace, debug, and improve Al agents using LangSmith.
- Deliver an end-to-end capstone: design, build, test, and present a practical Al agent prototype.
Prerequisites
- Own laptop with sufficient RAM and disk space, preinstalled with python, Node.js, Git, and Docker Desktop for the hands-on labs.
- Some signups for online tools will be required, using a Google account.
- Basic understanding of LLM concepts (prompting, tokens, limitations) or completion of an LLM fundamentals course.
- Working knowledge of Python (variables, functions, JSON); familiarity with APls is helpful.
- Comfort with practical automation thinking - breaking a task into steps, defining inputs/outputs, and handling errors.
Target Audience
- IT professionals and automation consultants transitioning from prompt engineering to building agent architectures.
- Enterprise software engineers building tool-integrated and multi-agent applications.
- Technical leads and system integrators implementing low-code and code driven Al automation.
- Professionals deploying Al agents across enterprise, factory, and development pipelines.
- Developers who want hands-on experience with MCP, A2A, Semantic Kernel, LangGraph, and LangSmith.
Training Outline
- Al Agent Concepts and Architecture
- Designing Agent Workflows
- Building Agents with No-Code and Low-Code Tools
- MCP for ALAgent Tool Integration
- AZA for Agent-to-Agent Communication
- Building Agents with Semantic Kerne!
- Introduction to LangGraph for Agent Workflows
- Evaluating and Improving AI Agents with LangSmith
- Capstone - Building a Practical Al Agent