Agents & Arbiters - An Adventurer’s Guide to Multi-Agent Collaboration with LangGraph.js

Building highly interactive systems with conventional coding is nigh impossible. You end up lost in a brittle maze of recursion and nested if statements. The more interactive you make it, the more complex your code gets, until it feels like being eaten by a grue—you know something's wrong, but you're just fumbling around in the dark. There's a better way. Instead of scripting interactions, we can give elements in our system their own intelligence. Multi-agent collaboration lets us create systems where entities are agents with their own perspectives and voices. Imagine an adventure game where the brass lantern, the white house, and the mailbox have something to say when responding to the player. Or consider a help desk where agents from billing, support, and legal weigh in to determine a course of action for a customer. In this session, we'll build a text-based adventure game—you know the one—that works like this. You'll meet the orchestration workflow—router, agents, arbiter, committer, and narrator—and discover how LangGraph.js coordinates the chaos. We'll shine our brass lantern over the code to see how it all works. And, to make your boss happy, we'll discuss how this same approach solves problems beyond gaming. When the adventure's over, you'll understand how to orchestrate agents to handle complex interactions. And you'll have a working example you can adapt for your own adventures—be they exploring the Great Underground Empire or building a customer service platform.