Session 10 Finale: Making AI Work

​Understanding and Building AI Agents

​We’ve spent Sprint26 Q3 building the pieces that make intelligent software possible: data pipelines, better search, AI tools, and workflows. Now it’s time to bring them together for the series finale.

​We’ll explore one of the most talked-about ideas in AI today: AI agents. What makes an AI system an agent? How can it decide what to do, use tools, remember information, and work through a task? We’ll break down the practical building blocks behind agents, then put them to work with n8n, a visual workflow platform, and tools developed throughout the quarter.

​We’ll look at how models, context, tools, and workflows combine to move AI from generating responses to taking useful action. We’ll also explore where people fit in, especially when an AI system is handling important or sensitive tasks.

​Then it’s time to build. Beginners can create an AI-powered workflow visually in n8n, connecting models, memory, tools, and external services. Developers can go further by connecting the workflow to the MCP tools and backend systems built during the Sprint.

​Together, we’ll create an agent that can understand a task, find relevant information, choose tools, work through multiple steps, and pause for human approval when needed before continuing. By the end, you’ll see how the quarter’s building blocks connect: data, context, search, tools, workflows, and agents.

​AI agents aren’t magic. They’re systems built from pieces we can understand, combine, and build ourselves. Join Infospace Meta in Nakuru to celebrate the Sprint26 Q3 finale and put those ideas into practice.