Parag Agrawal simply secured a $100 million Series A funding round for his new AI startup, Parallel Web Systems. Agrawal is the previous CEO of Twitter (now X).
His two-year-old firm, now valued at $740 million, is not constructing one other search engine for people. As a substitute, it’s tackling a a lot newer drawback: constructing internet search infrastructure designed particularly for AI brokers.
The transfer alerts that AI brokers are quickly turning into the online’s main customers. This implies the web’s core infrastructure might should be rebuilt to accommodate them.
To grasp the implications, I mentioned the information with Advertising AI Institute and SmarterX founder and CEO Paul Roetzer on Episode 180 of The Artificial Intelligence Show.
Betting Massive on AI Brokers
The funding spherical, co-led by Kleiner Perkins and Index Ventures, is notable not simply because it includes a high-profile founder however as a result of the quantity raised is so giant.
“That isn’t a typical elevate at a Sequence A,” Roetzer says. “That is a reasonably important quantity.”
He factors out that this sort of cash, from these top-tier buyers, signifies that enterprise capital corporations are actively “beginning to make some bets as to what the way forward for the web appears like.”
This $100 million funding is a transparent wager that the longer term is “agent-to-agent.”
“I believe everyone seems to be beginning to try to determine this out,” Roetzer says. “Firms like this are price listening to as a result of it is clearly kind of heading in that route of making an attempt to unravel for: ‘What does the following model of the web appear like?’ and ‘How does it have an effect on commerce and advertising and marketing and gross sales?’”
A Search Engine Tailor-made to AI Brokers
Parallel’s core premise is that AI techniques, like people, want entry to reside, up-to-date data from the online to carry out advanced duties. Enterprise clients are already utilizing its APIs to energy brokers that write software program code, analyze gross sales information, or assess insurance coverage dangers.
However conventional serps, which rank hyperlinks for people to click on, are inefficient for an AI agent.
Parallel’s system works otherwise. It returns “optimized content material, or tokens, designed to feed straight into an AI fashions’ context window.” The corporate says this improves accuracy, reduces AI hallucinations, and cuts operational prices.
A New Marketplace for Internet Content material
The cash raised will go towards product growth and buyer acquisition, but it surely’s additionally earmarked for a extra advanced problem: content material entry.
As AI internet scraping has change into extra frequent, many publishers and platforms have locked their content material behind paywalls and logins. Parallel’s answer is to make use of its capital to fund offers with on-line content material house owners and develop an open market mechanism.
This new financial mannequin would, in concept, incentivize publishers to make their content material accessible to AI techniques, making a secure and authorized information supply for the following technology of AI brokers. Though he didn’t present particulars on how this might work.
A Main Change for the Web
What this implies is that the web is shifting from a spot the place people browse to a spot the place autonomous AI brokers actively search, analyze, and act.
Roetzer says this transfer highlights the “continued want for us to be fascinated with what occurs when agent-to-agent turns into the norm on the internet.”
It’s a future that features brokers, not people, visiting your web site, and AI brokers interacting with chatbots.
If Parallel Internet Techniques makes this doable, a brand new type of web isn’t distant.
