Sierra, the enterprise AI company co-founded by former Salesforce CTO Bret Taylor, has acquired TakeOff, a startup building what it calls the leader in long-horizon AI agents. The deal was announced by Taylor on July 23, 2026, in a post on X, marking a significant consolidation move in the rapidly evolving market for autonomous AI agents designed to handle complex, multi-step enterprise tasks.
Financial terms of the acquisition were not disclosed. Neither Sierra nor TakeOff issued a separate press release, and no valuation or closing date was made public. But the transaction signals that established technology operators are moving to absorb nascent agent platforms before the market matures, a trend we are following as part of our broader AI industry coverage of enterprise artificial intelligence.
A Strategic Bet on Autonomous Agents
In his announcement, Taylor framed the acquisition as a way to bring TakeOff's agent platform into Sierra's portfolio, strengthening the company's ability to power autonomous, multi-step applications across enterprises. He was joined in the announcement by Clay Bavor, the former Google executive known for leading the company's AR and VR initiatives, including Daydream and Project Starline.
I'm incredibly excited to share that Sierra is acquiring TakeOff, the leader in long-horizon AI agents, Taylor wrote. Clay Bavor and I couldn't be more excited to welcome Aakash Thumaty and the entire TakeOff team. In less than a year, they've built something truly special.
The mention of less than a year underscores the speed at which the TakeOff team, led by founder Aakash Thumaty, built a platform capable of attracting an acquisition from a company of Sierra's stature. The rapid development timeline reflects both the maturity of the underlying large language model technology and the intense demand for agent platforms that go beyond simple chatbot interactions.
What Long-Horizon Agents Do
The distinction between a long-horizon agent and a conventional AI assistant is central to understanding why Sierra pursued this acquisition. Most commercial AI offerings today focus on short-term interactions — a chat assistant answering a single query, a coding tool generating a function, a search tool retrieving a document. These interactions are bounded in scope and duration.
Long-horizon agents, by contrast, are designed to plan and execute over extended sequences that can span minutes, hours, or even days. They maintain state across interactions, adapt to evolving goals, and coordinate across multiple tools or APIs without continuous human supervision. A long-horizon agent might be tasked with researching a market, drafting a strategy document, coordinating with team members, scheduling follow-ups, and revising based on feedback — all autonomously.
According to RuntimeWire, which reported on the deal, agents that can plan and execute over extended sequences remain rare in current commercial AI. Most offerings focus on short-term interactions, while turnkey platforms that let developers deploy long-horizon agents at scale remain elusive. TakeOff's claim of building a leader platform in under a year indicates rapid progress, though public technical details of the platform have not been disclosed.
Bret Taylor's Track Record
Taylor's involvement lends significant weight to the acquisition. His career spans some of the most influential products in consumer and enterprise technology. After co-founding FriendFeed, which was acquired by Facebook, he joined the social network and helped launch the Open Graph platform. He later served as CTO of Salesforce, overseeing the integration of Slack following its $27.7 billion acquisition.
Taylor also served as chairman of the board of OpenAI during a turbulent period that included the brief ouster and return of CEO Sam Altman in late 2023. He co-founded Sierra with Bavor to build conversational AI for enterprise customer service, and the company has become one of the more prominent startups in the enterprise AI agent space.
Clay Bavor's background in immersive technologies at Google — particularly Project Starline, a high-fidelity video communication system — suggests Sierra may also be exploring how agent capabilities intersect with interactive and spatial computing experiences.
The Consolidating Agent Market
The Sierra-TakeOff deal is part of a broader wave of consolidation in the AI agent market. As the technology has matured, larger companies and well-funded startups have been acquiring smaller teams with specialized agent expertise rather than building capabilities from scratch.
In 2026 alone, Cognition acquired Poke to bring AI personality to its Devin coding agent, Zoom acquired Common Room to bolster its AI revenue platform, and Stripe entered talks to acquire OpenRouter, the AI model marketplace, in a deal reportedly worth approximately $10 billion. Each transaction reflects a recognition that the next phase of AI value will come not from individual models but from the platforms and infrastructure that allow those models to operate autonomously in real-world workflows.
The long-horizon agent category specifically has drawn attention from researchers and investors who see it as the key to unlocking true AI autonomy. Demonstrations from DeepMind, OpenAI's multi-step reasoning models, and Anthropic's Claude series have shown fragments of this capability, but a platform that makes such agents reliable and deployable at scale has remained the missing piece.
What Comes Next
With TakeOff's team and technology integrated into Sierra, the company is positioning itself at the intersection of two of the most competitive areas in enterprise AI: customer-facing conversational agents and the autonomous back-end systems that can handle complex workflows without human intervention.
No timeline has been given for when TakeOff's technology will appear in Sierra's products, but the acquisition removes a potential competitor while adding capabilities that could differentiate Sierra from rivals in the crowded enterprise AI space.
The deal also raises a broader question about the long-horizon agent market: if platforms like TakeOff can be built in under a year and acquired before they gain widespread traction, how many independent agent platforms will survive? The answer may determine whether the next generation of AI applications is built on a handful of consolidated platforms or a diverse ecosystem of specialized tools.
For now, Sierra's bet is that owning the infrastructure for long-horizon agents is worth acquiring for, regardless of the price.
Stay Ahead of AI
The Sierra-TakeOff acquisition is the latest sign that the enterprise AI agent market is consolidating fast. For more on the deals, launches, and breakthroughs shaping the future of artificial intelligence, keep up with our latest AI developments.
Read more AI news →
