Parag Agrawal Returns to Tech With AI Startup Targeting Autonomous Web Research
Former Twitter CEO Parag Agrawal has stepped back into the technology scene with the launch of Parallel Web Systems Inc., a Palo Alto–based startup focused on helping artificial intelligence systems conduct research across the internet without human intervention.
Founded in 2023, the company has quickly raised $30 million from investors including Khosla Ventures, First Round Capital, and Index Ventures, and operates with a 25-member team.
How AI Agents Could Outperform Humans Online
Parallel’s platform is designed to let AI agents search, gather, and analyse information from multiple online sources with minimal delay.
According to Agrawal, the company’s flagship engine, Ultra8x, can perform deep research in roughly 30 minutes, while other engines deliver results in under a minute.
Early benchmarks suggest that Ultra8x outperforms leading AI models, including OpenAI’s GPT-5, on some research tasks.
Agrawal said,
“We already power millions of research tasks every day.”
He noted that some public firms are already using Parallel to automate workflows that were traditionally managed by humans.
From Twitter Leadership to AI Visionary
Agrawal rose through Twitter’s ranks as an engineer and later as chief technology officer under Jack Dorsey, before becoming CEO in November 2021.
His tenure ended abruptly in October 2022, when Elon Musk’s acquisition led to the dismissal of several top executives, including Agrawal.
The former CEO, along with others, later sued Musk over $128 million in unpaid severance, a case still pending resolution.
Following his exit, Agrawal immersed himself in AI research, writing code and studying papers.
He declined offers from struggling tech firms, stating he wanted to focus on AI rather than “clean up” other companies.
His first idea involved AI in healthcare, but he soon realised the internet itself would become dominated by autonomous AI agents.
He told Bloomberg,
“There’ll be more agents on the internet than there are humans around. You will probably deploy 50 agents on your behalf to be on the internet.”
AI Agents Seen as the Next Major Users of the Internet
Agrawal’s focus aligns with broader predictions from developers at Coinbase, who suggest autonomous AI agents could become Ethereum’s largest users.
By combining little-used web standards with blockchain protocols, AI agents could conduct transactions such as stablecoin payments autonomously, enabling applications ranging from self-driving taxis paying for fuel to automated content monetisation.
Parallel Aims to Serve Diverse Industries
Parallel’s technology is already being tested across various sectors.
AI-powered coding assistants can pull real-time code from repositories, retailers can track competitor products, and analysts can automate the compilation of data into reports.
Developers can integrate Parallel into their own platforms using APIs optimised for speed and scale.
With claims of outperforming leading AI models and backing from major venture investors, Parallel positions itself as a notable new player in Silicon Valley’s AI landscape.
For Agrawal, the startup represents a return not just to the tech industry, but to leading innovation in the rapidly evolving field of artificial intelligence.