Why This Series Exists
There is a real gap opening between companies that bolted a chat widget onto their site and companies building agents that retrieve knowledge, call tools, coordinate with other agents, and correct their own mistakes. The second group is harder to copy, and the patterns they use are mostly written down.
This series works through the 21 design patterns in Antonio Gulli's Agentic Design Patterns, one post at a time. Each post gives you the concept in plain terms, a code example you can run, and where the pattern earns its keep in a real deployment.
---
The Series
Foundations
- What Makes an AI System an Agent? The Foundation of Agentic Design. The five-step loop every agent follows, the four levels of agent complexity, and five hypotheses for where this is all heading.
Core Orchestration Patterns
- Prompt Chaining: Building Reliable AI Agent Workflows. Break a complex task into sequential steps, which is the foundational pattern everything else builds on.
- Routing and Parallelization: Scaling AI Agent Orchestration. How an agent chooses the right path at runtime and runs several steps at once.
- Reflection and Adaptation: How AI Agents Learn From Their Own Output. Self-correction feedback loops that turn first drafts into polished results.
External Integration Patterns
- Tool Use in AI Agents: Function Calling and Beyond. How an agent breaks out of the text bubble to call APIs and query databases and act in the world.
- Model Context Protocol (MCP): The New Standard for Agent-Tool Integration. The emerging protocol that standardizes how agents connect to tools and data sources.
- RAG for AI Agents: Retrieval-Augmented Generation Done Right. Ground your agents in factual, up-to-date knowledge with retrieval-augmented generation.
Planning & Reasoning
- AI Agent Planning: How Intelligent Systems Decide What to Do Next. Task decomposition, plan-and-execute strategies, and how agents turn goals into action sequences.
- Reasoning Techniques for AI Agents: Chain-of-Thought to Tree-of-Thought. The reasoning architectures that separate toy demos from production-grade agents.
Collaboration & Communication
- Multi-Agent Systems: Orchestrating Teams of AI Agents. How specialized agents work in concert to tackle objectives no single agent could handle.
- Agent-to-Agent Communication: How A2A Enables Agent Interoperability. The protocol layer that lets agents from different systems find each other and work together.
Memory & State
- Memory Management for AI Agents: Short-Term, Long-Term, and Beyond. How agents remember context across interactions and build persistent knowledge.
Reliability & Safety
- AI Guardrails and Safety: Building Trustworthy Agentic Systems. The safety patterns that make the difference between a demo and an enterprise deployment.
- Exception Handling and Human-in-the-Loop: Making AI Agents Resilient. Graceful failure recovery and knowing when to bring a human into the loop.
Operations & Optimization
- Monitoring AI Agents: Goal Setting, Evaluation, and Prioritization. How to set objectives for your agents, track how they perform, and decide what they work on next.
- Resource-Aware AI Agents: Optimization and Exploration Strategies. Cost-conscious agent design and intelligent exploration of solution spaces.
Perspectives & Resources
- Industry Leaders on Agentic AI: Perspectives from Google and Goldman Sachs. Saurabh Tiwary (VP, Google CloudAI) and Marco Argenti (CIO, Goldman Sachs) on the future of intelligent systems.
- The Agentic AI Toolkit: Frameworks, Environments, and CLI Agents. A practical tour of LangChain, CrewAI, Google ADK, AgentSpace, and the tools powering agent development today.
- The Definitive Glossary of Agentic AI: 100+ Terms Explained. Every term you are likely to meet in agentic AI, defined and cross-referenced.
---
About the Book
Agentic Design Patterns: A Hands-On Guide to Building Intelligent Systems by Antonio Gulli is published by Springer. It covers all 21 patterns with runnable code across three frameworks: LangChain, CrewAI, and Google ADK. Every royalty goes to Save the Children.
---
Space & Story helps brands get found and cited by AI; here is how that works.