After being ousted as Twitter CEO by Elon Musk in 2022, Parag Agrawal has returned to the tech scene with a bold new venture. He has launched Parallel Web Systems Inc., a Palo Alto-based cloud platform designed to help artificial intelligence systems conduct large-scale research across the web.

Founded in 2023, the startup has quickly gained traction, growing into a 25-member team and securing $30 million in funding from prominent investors such as Khosla Ventures, First Round Capital, and Index Ventures.

Powering AI Research at Scale

Agrawal revealed on LinkedIn that Parallel already powers “millions of research tasks every day” for both fast-growing AI startups and established enterprises. According to him, the platform enables AI-driven coding agents to search for technical documentation and debug issues, while one public company is using Parallel to automate traditionally human-driven workflows with results surpassing human-level accuracy.

Deep Research API: Outperforming Humans and AI Models

Parallel’s latest innovation is its Deep Research API, which Agrawal describes as the first of its kind to outperform not only human researchers but also advanced AI systems, including GPT-5, across some of the most complex benchmarks. The company believes this breakthrough signals the dawn of a new era where the web is no longer primarily for people but increasingly for machines.

Rethinking the Web for AI

In a blog post, Parallel outlined its vision for a “programmatic web” tailored for artificial intelligence. The company argues that while the open internet has served as humanity’s collective memory, the next phase of the web will be dominated by AI users. Unlike humans who browse casually or run short searches, AI systems may sift through entire databases, process vast amounts of information, and extract insights at massive scale.

However, current web business models—built around ads, clicks, paywalls, and gated APIs—are ill-suited for machine-driven use. Agrawal’s startup suggests that the internet must evolve into a framework designed for reasoning, computation, and verifiable sources, enabling AI to interact with knowledge more effectively.

A Vision Beyond Human Browsing

With Parallel, Agrawal is betting on a fundamental shift: that the internet, which was built for human attention, must now adapt to meet the needs of artificial intelligence. His venture positions itself as a pioneer in building the infrastructure for an AI-first web, hinting at how digital ecosystems may soon transform.