At Felicis, we’ve known for a while that agents will fundamentally reshape how software interacts with the world. While the first wave of AI agents focused on executing actions across APIs, the second wave of AI agents will interact with the web at a deeper level. It’s no longer speculative to suggest that the vast majority of web traffic in the near future will be non-human. As software becomes more autonomous, the browser emerges not just as an interface for humans, but as the execution environment for intelligent agents.

Enter Browser Use, the easiest way to connect your AI agents with the browser. It enables AI agents to interact with websites through a powerful yet easy-to-use browser automation interface.

Browser Use is creating one of the most important pieces of AI agent infrastructure: web interaction. Browser Use extracts HTML and elements to make it much easier for an agent to interact with a website and repeat the workflow as often as required. It supports multi-step workflows, persistent sessions, and even intelligent self-correction when page layouts shift. And unlike vision-based systems that require expensive per-instance inference, Browser Use agents can deterministically re-run workflows— improving reliability, speed, and cost-efficiency.

The browser is where real work happens: e-commerce transactions, document handling, form filling, and workflows. These systems are rarely API-first, and even when APIs exist, they often fail to capture the full functionality or complexity of the UI. Browser Use enables AI agents to interact with the web the way humans do: visually, contextually, and reliably.

Browser Use founders Magnus Muller and Gregor Zunic are builders first and foremost. They met at a hacker house at ETH Zurich and experienced firsthand the limitations of existing browser automation stacks. Rather than settle, they built what they wished existed—and in doing so, created a platform that has already seen explosive adoption. Spending time with them, we immediately felt their relentless energy, fueled by a deep obsession with what they’re building.

Browser Use launched its open source project in November 2024, and it quickly became one of the fastest-growing projects in the developer ecosystem, skyrocketing to 46K GitHub stars. It’s super easy to install so check it out! Just pip install browser-use to get started or sign up for their managed service. 

We are excited to announce that Felicis led Browser Use’s $17M seed round. We are joined by A Capital, SV Angel, Nexus Venture Partners, and Paul Graham among others. We are incredibly excited to support Magnus, Gergor, and the entire team as they build the definitive open platform for web browsing AI agents.

If you’re excited about building at the frontier of AI agent software, we encourage you to get involved. You can follow Browser Use on Twitter and join their Discord. They are actively hiring so check out opportunities here!