Descriptions
Agentic AI, Agentic AI represents a new way of building software that uses large language models (LLMs) to complete some or all steps of complex tasks. Instead of generating a single response to a prompt, agentic workflows can plan multi-step processes, execute them iteratively, and improve their outputs through reflection and tool use. In this course, you will learn to build sophisticated agentic AI systems from the ground up using Python. You will explore 4 core design patterns: reflection, in which AI critiques and revises its own work; tool use, which connects AI to databases, APIs, and external services; planning, which breaks complex tasks into executable steps; and multi-agent workflows, in which specialized AI agents collaborate to handle different parts of a task. You will learn how to decompose business processes into agentic workflows and build practical solutions such as research agents, email assistants, and automated task pipelines, while understanding how to customize workflows before relying on frameworks.
The course also emphasizes evaluating, debugging, and optimizing agentic systems for reliable production use. You will learn to create testing frameworks, analyze errors and execution traces, evaluate both end-to-end workflows and individual components, and use metrics such as precision, recall, and F1 where appropriate. You will explore how to improve performance through prompt design, model selection, hyperparameter tuning, and optimization of latency and cost. In the advanced modules, you will learn how LLMs can generate executable plans using structured JSON or code, orchestrate tools and data sources, and coordinate multiple specialized agents through different communication patterns. Practical examples and programming assignments will help you understand the trade-offs between flexibility, security, predictability, and autonomy, and develop the skills needed to build, evaluate, and deploy effective agentic AI workflows.
What you’ll learn
- Understand the fundamentals of agentic AI workflows and identify when they are more suitable than single-prompt approaches.
- Implement the reflection design pattern to evaluate, revise, and improve LLM-generated outputs using explicit criteria and external feedback.
- Connect LLMs to external tools, including APIs, databases, and code execution, using tool calling and function calling.
- Use Python to build agentic workflows and decompose complex tasks into smaller, executable steps.
- Design and implement planning workflows in which LLMs generate executable plans using structured JSON or code.
- Build multi-agent systems in which specialized agents collaborate through linear, hierarchical, or other communication patterns.
- Evaluate and debug agentic systems through systematic error analysis, execution traces, component-level evaluations, and metrics such as precision, recall, and F1.
- Optimize agentic workflows for quality, reliability, latency, and cost, and prepare them for production deployment.
Who this course is for
- Python developers who want to build practical agentic AI applications and workflows.
- AI and software engineers interested in implementing LLM-based systems that can plan, use tools, reflect, and collaborate across multiple agents.
- Developers who want to integrate language models with APIs, databases, web services, and code execution environments.
- Technical professionals who need to evaluate, debug, test, and optimize AI applications for reliable production deployment.
- Learners with intermediate Python skills and a basic understanding of large language models and APIs who want to develop more advanced AI systems.
Specificatoin of Agentic AI
- Publisher : Coursera
- Teacher : Andrew Ng
- Language : English
- Level : Intermediate
- Number of Course : 5
- Duration : 2 weeks to complete at 10 hours a week
Content of Agentic AI

Requirements
- Intermediate Python skills and a basic understanding of large language models and APIs
Pictures

Sample Clip
Installation Guide
Extract the files and watch with your favorite player
Subtitle : English
Quality: 720p
Download Links
Password file(s): www.downloadly.ir
File size
1.31 GB


