Concept launches agents for data analysis and task automation

At the “Make With Notion” event on Thursday, the company announced the launch of its first AI agent. Agents use all user concept pages and databases as context to automatically generate notes and analysis for meetings, competitor rating reports, and feedback landing pages.

According to the productivity platform, agents can create pages and databases and update them with new data, properties or views. Users can also trigger concept agents from external platforms linked to the service. For example, you can ask a concept agent to create bug tracking dashboards from a variety of sources, such as Slack, Email, Google Drive, etc.

The newly announced agents are based on concept AI, an existing feature that allows you to search or summarize content. However, new agents can use the power of agent AI to tackle more complex multi-step tasks. The company said the current version of the agent can perform tasks that run for up to 20 minutes on hundreds of pages.

Users can set up a Profile page to instruct agents to follow instructions for references to sources, output styles, and where to update tasks and final results. You can also ask agents to “remember” key points when people use them. These memories are saved on the profile page, where users can edit them.

In the demo video, the company cited examples of agents who can provide feedback on landing pages, update them, create restaurant trackers, create analytics from memo meetings, and prepare competitive analysis reports.

At this time, you will need to manually trigger these actions. However, the concept stated that the ability to create customized agents that work with schedules or triggers will soon be emerging. The company will also release a template library for agents, allowing you to select a ready-made prompt that suits your task.

Over the past two years, Notion has released an enterprise search to retrieve information from calendar apps, Gmail clients, meeting note takers, and a variety of sources. These are the features that gave the company enough context building blocks to create automation. Other enterprise knowledge and productivity platforms, such as Salesforce, Fireflies, and Read AI, have launched their own agents to extract and update information.