Former Twitter CEO Parag Agrawal closed a $100 million Series B for Parallel Web Systems, the web infrastructure startup he founded in early 2024. Sequoia Capital led the round, with Kleiner Perkins, Index Ventures, and Khosla Ventures participating. The raise values the company at $2 billion post-money and brings total funding to $200 million, following a $100 million round in November 2025.
The Product
Parallel builds specialized APIs that let AI agents interact with the web programmatically rather than through browser interfaces designed for humans. The platform includes four core products: search, task execution, data extraction, and web monitoring. All four run against a proprietary “machine-optimized” web index built for retrieval by autonomous systems rather than human queries.
Agrawal told the Wall Street Journal that agents will “use the web a lot more than humans” and require fundamentally different infrastructure to access it. The specific use cases he pointed to, according to SiliconANGLE, include insurance claims processing and sorting through government contracts, tasks where agents can move faster than humans if they have programmatic access to web data rather than a rendered browser view.
Traction and Competition
Parallel has accumulated over 100,000 developer users since launch, including AI startups like Harvey AI and unnamed enterprise customers. Harvey co-founder Gabe Pereyra told the WSJ that giving agents access to Google Search is insufficient. They need “more granular control over which websites they should be accessing,” which is what Parallel provides.
Sequoia partner Andrew Reed framed the investment around long-horizon agents that operate continuously in the background. “One of the things that is a core shared function amongst all of these long-horizon agents is the ability to use the web,” Reed told SiliconANGLE.
The company is not alone in the space. Tavily (acquired by Nebius in February) and Exa Labs both build competing infrastructure for agent web access. The $2 billion valuation signals investor conviction that agent-native web infrastructure is a category worth funding aggressively, not a feature that existing search providers will subsume.
The Bet on Agent-Native Data Architecture
Agrawal plans to use the capital to build out sales and marketing operations and accelerate R&D, according to the WSJ. The timing aligns with a broader shift: as agents move from demo to production, they expose gaps in existing web infrastructure. Browser automation is fragile. Search APIs return results optimized for human reading patterns. Parallel’s thesis is that the entire data access layer needs to be rebuilt for machine consumers, and that whoever owns that layer captures a toll on every agent workflow that touches the open web.