VC is doing Roll-up, the AI Infra grows at 20x in ARR
Deploy and scale AI using open-source models
As AI foundation models become increasingly mature, venture capitalists are no longer content with traditional investment models—they’re now exploring new strategies.
Last month, Sarah Tavel, a partner at Benchmark, announced she would transition from a General Partner to a Venture Partner, aiming to dive deeper into AI through hands-on coding and writing. She’s taking a “Vibe Coding” approach to explore AI firsthand.
AI isn’t just another technology wave – it’s a transformational force that will reshape every corner of our lives. As I've watched the space evolve, I've increasingly realized that for me to best support founders building in AI, I need more space. More space to go AI-native by fully immersing myself in AI tools at the edge, more space to reflect on the last ten years and try to understand what the next ten might bring, and maybe even (don't hold me to this!) more space to think through it all by writing.
Now, more and more VCs are getting directly involved in incubation and acquisitions. They’re leveraging AI and their ability to integrate resources to transform traditional industries through automation.
This strategy—acquiring multiple similar businesses and using AI to upgrade their operations, improve efficiency and customer experience, and achieve economies of scale—is referred to as the AI-powered Roll-up strategy.
What does this strategy look like in practice?
This private-equity-style strategy could offer an unexpected advantage to the many AI startups backed by VCs. By pairing legacy businesses with cutting-edge AI, VCs can give these startups immediate access to sizable, well-established customer bases.
Acquiring several mature but traditional businesses to build a network.
Using AI automation tools to optimize operations (e.g., AI-powered customer service, intelligent financial processing, data-driven decision support).
Leveraging AI’s predictive capabilities to enhance customer management and market expansion.
Reducing operational costs and improving profitability through technical integration.
The targets are often labor-intensive traditional sectors such as call centers (e.g., the AI customer support startup I mentioned before that hit $90M ARR in just 15 months), accounting firms, and similar.
Several major players are now actively exploring this direction, including Thrive Capital, General Catalyst, Khosla Ventures, and even solo VC Elad Gil.
General Catalyst is perhaps the most aggressive, positioning itself as a global investment and transformation firm. It has already backed about seven companies using this strategy.
One such company, Long Lake, was founded just last year and has raised $670 million, although it has yet to launch publicly. It is a long-term holding company partnering with top U.S. service sector operators to apply cutting-edge technology.
Khosla Ventures, traditionally an early-stage investor, is also exploring this path. General Partner Samir Kaul said that this strategy is especially useful when AI startups struggle to acquire customers on their own.
With the rapid rate of change in AI, the number of startups pouring into the market, and the historically long sales cycles involved in selling to enterprises, such difficulties apply to many AI startups.
The AI Infra startup grows at 20x in ARR
The explosive growth of AI application have led to a huge demand for inferencing AI infrastructure. Then an AI Infra startup offering GenAI Platform as a Service is growing at 20x yoy, as its ARR is said to hit $130M in 2025.
What it is doing is to help developers and businesses efficiently connect to the right AI capabilities—focusing on usability, quality, cost-efficiency, and security. It aims to democratize GenAI infrastructure by shortening training and inference times, enabling the rapid growth of AI applications.