Asana was battered by the AI boom. Now it’s betting its future on humans and agents working together.

On Thursday, Asana announced that it had acquired Stack AI, a no-code AI agent builder, for $75 million—its first acquisition in 18 years—timed to land alongside a first-quarter earnings beat that sent the company’s shares up more than 13%.

The acquisition is aimed at repositioning Asana as a platform for managing AI agents alongside human workers, at a moment when the company’s core business model is under intense pressure to adapt for the AI age. Asana has fallen victim to deep market anxiety regarding the future of seat-based SaaS models in an era of agentic AI. AI can increasingly do the work that the SaaS product itself was built to do, sparking investor concerns about the future need of such services. Companies like Asana have also historically grown by charging per employee seat, where more headcount meant more revenue. AI agents, which can handle work that previously required multiple human users, upend that business model. 

Fears around a potential SaaSpocalypse erased more than $1 trillion in SaaS market capitalization in February alone, as investors began pricing in a structural contraction across the sector. Over a tumultuous year, Asana’s stock has fallen from $19 at its 52-week high to a low of $5.38. Thursday’s deal was partly meant to answer the question of what Asana actually is in a world where AI does a lot of what work-management software was built for.

Asana CEO Dan Rogers, who is less than a year into the role following co-founder Dustin Moskovitz’s departure, is pitching Asana’s future as the coordination layer that makes human-agent collaboration actually work at enterprise scale. He told Fortune that as AI agents proliferate across enterprises, the coordination problem just gets harder. In two or three years, he said, most workers will have agents augmenting and supercharging the way they work, making the question of how humans and agents stay aligned more urgent. 

“The coordination and collaboration challenge moves from human to human to human to agent,” he said. “Asana is becoming the operating system for human-agent teams.”

The Stack AI acquisition is aimed at accelerating the company’s shift to managing these enterprise agents. Built as a no-code platform for deploying agents across enterprise systems, the startup runs AI agents that can complete complex workflows end-to-end across multiple systems—such as employee onboarding or taking in marketing content, performing quality control checks and then publishing it using CMS software. Rogers said this was also the eventual plan for Asana’s own AI products, and the acquisition is a way to accelerate that development.

“If you looked at the roadmap of the things they were building and the roadmap of the things we were planning on building, it’s a perfect overlap,” he said, adding that he expects full integration within two to three months.

Stack AI’s co-founders, Toni Rosinol and Bernard Aceituno, will join Asana along with the company’s full team of around 55 people. Stack AI had raised just under $20 million prior to the acquisition, including a $16 million Series A from investors including Gradient, Epakon Capital, and Vercel CEO Guillermo Rauch.

Asana also announced its earnings on Thursday. Revenue for Q1 came in at $205.1 million, up 9.5% year over year and above the high end of guidance. The company is still loss-making on a net basis, but new AI products like AI Studio and AI Teammates, both launched within the past year, now account for more than 17% of new ARR, according to Rogers, and the number of customers spending more than $100,000 annually on AI Studio nearly doubled during the quarter.

While Asana’s AI tools appear to be popular with users, the same cross-system agent orchestration that Stack AI brings is also being built by companies like Salesforce and ServiceNow. Rogers argues that Asana’s horizontal footprint within companies—where it is already embedded across marketing, IT, operations, and planning in large enterprises—gives it a natural coordination role that larger rivals cannot easily replicate. Still, the road back will not be without tough competition.