Notion’s Sarah Sachs wants AI agents to work for everyone

Every few weeks, Sarah Sachs and her team swap in the best-performing AI models across the productivity and collaboration app Notion, a rotation that plays out invisibly for more than 100 million users. “I’ve always believed the real work begins when AI comes out of the chatbot,” says Sachs, who leads artificial intelligence modeling at the productivity-focused company. Sachs’s teams are building agents that can triage inboxes, help run meetings, and manage projects across tools like Slack, Git Hub, and Google Drive. Early testers created more than 21,000 Custom Agents before Notion even launched that capability publicly in February, while Notion employees created 2,800 agents of their own. Sachs says Notion’s AI has evolved from simple rewriting tools to retrieval-based Q&A and now to what she calls “governed AI teammates” that operate with their own permissions and audit trails. When reasoning models improved enough to complete sequential tasks by themselves, her team rebuilt Notion’s entire AI architecture from scratch to take advantage. Sachs is also refreshingly vocal about ensuring that AI doesn’t become a tool reserved for deep-pocketed companies. She argues that applied AI companies should negotiate aggressively with frontier model providers and invest in open-weight alternatives, where the parameters of a trained neural network are publicly shared. “We feel really strongly that we’re like the Switzerland of AI models,” she says, because she wants to make sure “the Fortune 5,000,000, not just the Fortune 500, can afford to use agentic AI for valuable work.” This profile is part of Fast Company’s AI 20 for 2026, our roundup spotlighting 20 of AI’s most influential technologists, entrepreneurs, corporate leaders, and creative thinkers.