CopilotKit raises $27 million to assist developers in implementing app-native AI agents

CopilotKit is aiming to change how AI works inside applications, moving beyond traditional chatbot-style interfaces. The startup recently raised $27 million in a Series A funding round led by Glilot Capital, NFX, and SignalFire.

Today, many apps use AI as simple chatbots where users type commands and receive long text responses. However, CopilotKit founders Atai Barkai and Uli Barkai argue this approach is inefficient and clunky. Instead, they believe AI agents should be embedded directly into applications, where they can understand user actions, perform tasks, and display interactive interfaces.

At the core of this vision is CopilotKit’s open-source AG-UI protocol. This framework allows AI agents to connect with app interfaces, enabling features like real-time interaction, tool usage, and shared state between the user and AI. Rather than returning long paragraphs, AI agents can present information visually — such as charts, dashboards, or interactive elements — tailored to the app’s design.

For example, if a user asks for revenue insights, the AI could generate a pie chart instead of text, making the experience faster and more intuitive. Developers also have control over how much freedom the AI has in designing these interfaces, from strict layouts to flexible components.

The company is building an enterprise toolkit on top of AG-UI, offering features like self-hosting and advanced support. This comes as adoption grows rapidly, with major tech companies like Google, Microsoft, Amazon, and Oracle supporting related standards.

CopilotKit says its tools are already used by large enterprises including Deutsche Telekom, DocuSign, Cisco, and S&P Global.

Despite competition from platforms like Vercel and OpenAI, CopilotKit is positioning itself as a flexible, enterprise-focused solution for integrating smarter AI agents directly into apps.