Miro CEO: Redesign Work Around AI, Not Bolt It On
Miro co-founder and CEO Andrey Khusid stated in Turin that companies should not simply add AI to their existing teams and processes but instead redesign workflows from the ground up to fully leverage its potential. He made these remarks a month after Bending Spoons agreed to acquire Miro for $1.355 billion.
"One of the big challenges is how not to put AI on top of the existing teams or operational systems, but actually... design the workflow and then only add humans to add value on top of that workflow," Khusid emphasized.
This session, titled "Thinking Together in the Age of AI," was moderated by Charlie Perreau from Les Echos.
Khusid noted that while individuals are increasingly using AI tools like Claude or ChatGPT daily, transforming entire companies' operations is a greater challenge. He explained that companies operate on complex workflows where information flows from one person to another, each holding only a fragment of the context. The initial step is consolidating this context so teams can work with it effectively.
Implementing AI across an existing operational model will not yield the expected gains, according to Khusid. Miro addressed this by setting up parallel teams for specific tasks utilizing AI and then structuring processes around them before scaling them company-wide.
He also highlighted a potential downside of AI: while it provides quieter individuals with a more effective method to share their ideas, it generates far more content. If everyone feeds AI outputs into their teams, it can overwhelm them and slow down operations. Khusid suggested accountability as a solution, emphasizing that each person must take ownership of every piece they contribute.
He compared traditional organizational structures, built around people and hierarchy, to newer companies where central intelligence allows swift decision-making and faster operation. Established companies, he argued, need to rebuild their intelligence infrastructure and reevaluate the roles they require.
According to Khusid, Miro, which is 15 years old, has already incorporated AI assistance into much of its code and merged systems to enhance AI context. While there has been little disruptive change in models recently, he believes it's clear what AI can achieve over the next three years, allowing Miro to focus on building around that rather than chasing every new release.