Former Twitter CEO Parag Agrawal has introduced Parallel Web Systems Inc., a cloud platform built to help artificial intelligence systems conduct large-scale online research.
Mr Agrawal, who was fired by billionaire Elon Musk after he took over Twitter (now known as X) in 2022, founded Parallel in 2023 and has since assembled a 25-member team in Palo Alto.
The company has already secured $30 million in funding from marquee investors, including Khosla Ventures, First Round Capital, and Index Ventures.
"We already power millions of research tasks every day, across ambitious startups and public enterprises," Mr Agarwal wrote on LinkedIn.
"Some of the fastest growing AI companies use Parallel to bring web intelligence directly into their platform and agents. A public company automates traditionally-human workflows exceeding human-level accuracy with Parallel. Coding agents rely on our search to find docs and debug issues," he added.
Parallel has also announced the launch of its Deep Research API, which Mr Agrawal says is the first to outperform both humans and leading models, including GPT-5, on two of the most challenging benchmarks.
Inside Parallel
Parag Agrawal's new startup, Parallel, is betting on the idea that the internet, as it exists today, was built for people but the next era belongs to artificial intelligence.
In a blog post, the company said:
- The web is humanity's memory. The open internet has enabled publishing, learning, and collaboration at scale. It is the foundation on which modern AI was trained.
- AI is becoming the web's primary user. Unlike humans, who browse a few pages or run short searches, AIs might comb through entire databases, spend hours processing information, or pull facts instantly at massive scale.
- Old business models won't work. Today's web relies on human attention like clicks, ads, paywalls, and gated APIs. But these models aren't designed for machine use, and they risk locking valuable knowledge behind silos.
- A "Programmatic Web" for AIs. Parallel argues the internet must evolve into a system designed for machines, one that supports reasoning, computation, and verified sources.
Key Principles Of Parallel
The company outlines key principles of this new web:
- Unified infrastructure combining data, compute, and reasoning, producing insights and actions instead of just static documents.
- Declarative interfaces where AIs say what they need, and the system decides how to get it.
- Transparent attribution so every source is credited, and contributions can be measured.
- Open markets where value is rewarded economically, ensuring openness thrives not just by goodwill but by financial incentives.
Parallel frames this as either the web adapting to serve its "second user" (AI) or it risks breaking apart.
How Parallel Will Work
Parallel is creating a version of the internet built for AI. Instead of humans clicking and searching, AIs will be able to ask for information directly, and Parallel's system will gather, process, and organise it. It will also give credit to sources and reward those who contribute. In short, Parallel makes it easier and fairer for AIs to use the web to find answers and create insights.
Parallel's mission is about "building for abundance".
