When Andrey Khusid cofounded Miro in 2011, the idea was simple: bring a whiteboard into the browser, and let people collaborate visually, not just with text. Now that digital canvas is evolving into what the CEO calls an “AI Innovation Workspace.” More than 100 million people use Miro these days, so the company’s ventures into AI are quickly reaching more than 250,000 organizations, including GitHub, Prudential, and Cisco.
To serve those Fortune 500 companies, Miro now offers a platform for collaborative AI workflows with Sidekicks that work alongside teams on the canvas, and tools for turning rough sketches into clickable prototypes. The company, which sported a $17.5 billion valuation when it raised $400 million in 2022, recently acquired Butter, a workshop facilitation platform, to tackle what Khusid sees as broken meetings.
Fast Company spoke with Khusid about why he believes AI’s real value lies in teams rather than individuals, how Miro can compete in an increasingly crowded software arena, and what he learned from scaling the company from 200 to 1,600 employees.
The conversation has been edited for length and clarity.
You’ve said AI’s biggest opportunity lies in teamwork, not individual productivity. Why is focusing on solo work a mistake?
When you work in a company, the company moves not with the speed of every individual. The company moves with the speed of the outcome that those individuals produce together. If I boost my individual productivity but I still wait two weeks between my milestone and the next person’s milestone, it doesn’t matter if I accomplish the task in one hour instead of one week.
Before, you would do a workshop, then someone needs to summarize action items—that takes a week. Someone comes back, the group forgot what was discussed, you chase people—another week or two. Then someone breaks it down into tasks, someone builds a prototype. What AI allows you to do is see the final output in the first workshop. You brainstorm ideas, turn them into a project brief, a project plan, a prototype—and you see it immediately. You work backwards and iterate on the inputs. That’s a completely different game.
