Vision has recently transformed its workspace into a hub for AI agents

Production software maker Notion is entering the age of the agent.

In a live product announcement Wednesday, the company, best known for its collaborative note-taking app, introduced a new platform for developers that expands the capabilities of its custom AI agents, connects externally, and allows teams to build automated multi-step workflows that can pull data from any database.

By building an orchestration layer — a system that connects AI operations across multiple tools and data sources — Vision positions itself as more than a mesh of AI features and instead as a hub where people and agents can interact across tools and databases.

In February, Notion first introduced Custom Agents – AI teammates that handle repetitive tasks, such as answering frequently asked questions, compiling status updates, and automating workflows. Since then, Notion customers have created more than a million agents, the company said.

However, these agents had limitations. They could not communicate with external data or use custom logic. The external agents used by the companies also did not have a way to communicate with the Notion workspace. Teams had to solve these problems by using third-party automation platforms or writing their own scripts that worked with their infrastructure.

“It’s true that historically, Notion has not been a very developer-centric platform,” said Ivan Zhao, Notion’s founder and CEO, during the live stream. “But things change.”

Now, Notion will allow teams to release their own custom code. With its new Workers, Notion’s cloud-based environment for running custom code, customers can write their thinking and send it to a secure sandbox (an isolated place that keeps the code from interfering with other systems). This allows teams to do things like sync their data to Notion, build custom tools, and trigger work with webhooks — which are automatic signals that trigger actions when something happens in another application — without needing to rely on external infrastructure.

You don’t even need to write code. The company claims that your favorite AI coding agent can do it for you.

Agents will use the same credit system as Custom Agents, but Notion is making this free until August, so developers can try it out.

Synchronizing external data sources is also part of the Notion Developer Platform. Powered by Workers, the database synchronization feature can pull data from any database via an API. That means you can access data from places like Salesforce, Zendesk, Postgres, and others within your Notion database — and keep the data up-to-date.

Zhao noted that this means Notion users can now “use your Notion database as just a canvas to power both your workflows and your agents.”

Employees can also build agent tools with custom logic, for those times when communicating with a third party via MCP — short for Model Context Protocol, the emerging standard that allows AI tools to connect to external data and services — isn’t enough.

Another addition allows Notion users to chat directly with the external AI agents they use, assign them tasks, and track their progress, just as if they were one of Notion’s own custom agents. At launch, Notion says Claude Code, Cursor, Codex, and Decagon are the supported interactive agents, but plans to add more.

There’s an External Agent API, too, if teams want to connect their internal agents with Notion, such as those they’ve built specifically for their company’s needs.

Developers and agents interact with Notion’s Developer Platform through the Notion CLI, a developer command tool, available in the company’s Business and Enterprise Plans.

The Developer Platform represents a shift in Notion’s strategy as it becomes a programmable platform rather than just an application, making it competitive with other workflow automation platforms. As businesses increasingly look to automate information work and build internal AI systems, a platform that integrates agents, custom code, and live data in one place is starting to look less like a production application and more like core infrastructure.

It also follows a broader trend among AI companies, which have been moving beyond the AI ​​chatbot to provide actionable agent tools across software platforms.

“Any data, any tool, any agent — that’s the big picture of the Notion Developer Platform,” Zhao said.