Former Twitter CEO Parag Agrawal Unveils AI Startup Parallel Web Systems The firm introduced its flagship product, the Deep Research API, which is aimed at enabling AI agents to carry out advanced real time web research.
You're reading Entrepreneur India, an international franchise of Entrepreneur Media.
Former Twitter CEO Parag Agrawal has entered the AI sector with the launch of Parallel Web Systems, a company headquartered in Palo Alto. The firm introduced its flagship product, the Deep Research API, which is aimed at enabling AI agents to carry out advanced real time web research.
Agrawal, who led Twitter until its takeover by Elon Musk in 2022, has re-emerged in the technology landscape with this new venture. Reports indicate that Parallel Web Systems has secured funding of around USD 30 million from investors such as Khosla Ventures, Index Ventures and First Round Capital.
The company stated that the Deep Research API is already managing millions of research tasks daily. These include functions such as discovering documents, debugging code and automating workflows. Agrawal said the system is designed to reach levels of accuracy beyond human capability in certain specialised tasks.
The product incorporates eight research engines, each designed for specific computational activities. These engines cover areas like long form synthesis, cross disciplinary analysis and knowledge retrieval. The tool is positioned for developers and organisations that wish to embed advanced research functions into their applications.
According to company claims, the Deep Research API has outperformed leading models including OpenAI's GPT 5 in selected benchmarks related to online research. Independent reviews are still limited, but early demonstrations suggest that the platform can manage complex, multi step queries.
Industry analysts see the development as timely, given the rising need for AI systems that access and interpret live data instead of relying only on static training sets. Potential applications include market intelligence, legal analysis and scientific research, with developers expected to use the system in autonomous agents and decision support solutions.
