Building AI Coding Agents and Agentic Design Patterns

As AI transforms software engineering, the real power lies in building systems that can plan, reason, and execute tasks autonomously. Moving beyond simple chat prompts to structured agentic workflows is the next critical step for modern developers. This text-based course guides you through the foundational concepts and architectural patterns of AI coding agents. You will transition from writing basic prompts to designing robust, multi-agent systems that can automate complex coding workflows, debug errors, and orchestrate tasks systematically. What you'll learn: - Understand the core concepts of agentic reasoning, memory, and tool calling. - Implement key design patterns including prompt chaining, parallelization, and routing. - Design orchestrator-worker architectures to break down complex development tasks. - Apply human-in-the-loop patterns to ensure safety and quality control in automated workflows. - Explore modern evaluation techniques to test and optimize agent performance. You will start with essential terminology and basic agent structures before progressing to step-by-step written breakdowns of complex design patterns. Through clear explanations and practical code walkthroughs, you will learn how to coordinate multiple agents to solve real-world programming problems. This course is designed for software developers, engineers, and tech enthusiasts who want to build agentic systems, requiring only basic programming knowledge to get started. Start reading today to master the architectural patterns shaping the future of software development.