Parallel Web Systems, the AI startup founded by former Twitter CEO Parag Agrawal, is reported to have secured $100 million in a Series A round co-led by Kleiner Perkins and Index Ventures.
The company is building a new layer of web infrastructure designed for AI agents to search and interact with live, accurate data, Cyber News Centre reported.
The funding round also had participation from Spark Capital and existing investors including Khosla Ventures, valuing the company at $740 million.
The platform addresses the limitation of large language models that rely on static, outdated datasets, which can lead to inaccurate or “hallucinated” outputs.
By creating a new layer of web plumbing, Parallel enables AI agents to access and process fresh, reliable online information for tasks such as software development, sales data analysis, and insurance risk assessment, according to Reuters.
Agrawal told Reuters its enterprise customers use Parallel to power AI agents that write software code, analyze customer data for sales teams and assess risk for insurance underwriting — areas where high-quality web data, alongside internal systems, is critical.
“How many jobs are there where we could turn off web access and ask you to do the same job fully?” Agrawal was quoted as saying. “You can’t deprive an M&A lawyer from not being able to use the web, so why would you deprive their agents?”
He said he believed the startup’s technology was superior to built-in web search functions offered by AI-model providers.
Unlike traditional search engines that rank links for humans to click, Parallel’s system returns optimized content, or “tokens,” designed to feed directly into an AI model’s context window.
READ: Startup founded by former DeepMind researchers Reflection AI raises $2 billion (
The company says this approach improves accuracy, reduces “hallucinations,” or false information, and cuts operational costs for customers.
Agrawal said the new capital will allow Parallel to “go all in” on product development and customer acquisition.
Agrawal said Parallel plans to develop an “open market mechanism” — a new economic model to incentivize publishers to keep content accessible to AI systems, though he did not provide details.
Kleiner Perkins partner Mamoon Hamid will join the company’s board. In a statement, Hamid compared the investment to his firm’s early backing of Google, noting, “26 years later, Parallel is indexing the web for its ‘next’ user: AI agents.”
The company’s first products, including a web search API for agents, were launched in August 2025.