Case Study: Cedar builds multi-agent AI copilots with Mastra

How Cedar Built Multi-Agent AI Copilots with Mastra's Modular Architecture

Cedar, a startup founded by Jesse Li and Isabelle Ilyia, was exploring AI onboarding agents but faced challenges in creating a generalized system. Their main hurdle was coordinating specialized agents for chat, DOM manipulation, and workflow execution seamlessly and at scale, rather than focusing on their core product. This required them to solve complex infrastructure problems like agent orchestration and streaming from scratch.

Mastra provided the solution with its modular, open-source TypeScript framework. Cedar used Mastra's architecture to build a sophisticated multi-agent system, or "agent army," where specialized agents are orchestrated to provide contextual, step-by-step guidance. This allowed Cedar to achieve transformative development velocity and focus on their business and product, as Mastra handled the underlying infrastructure. The result is a powerful onboarding experience where customers see setup tasks performed in real-time.

Cedar

Isabelle Ilyia

Chief Technology Officer