Best Internet Providers in Valdosta, Georgia

AI-powered marketing automation startup Conversion, founded five years ago by two UC Berkeley dropouts, has raised a $28 million Series A led by Abstract, with participation from True Ventures and HOF Capital.

The company’s founding story sounds like it could have been an episode of the HBO show “Silicon Valley.”

The story begins all the way back when co-founder and CEO Neil Tewari, now 24, was in high school. He got busted one day watching a TechCrunch Disrupt livestream during class, was sent to the principal’s office, and had to stay late.

Afraid to call his parents and tell them why they had to pick him up, he instead called one of their close friends. On the drive home, Tewari explained to the friend what got him in trouble. “I told him I had this interest [in entrepreneurship], and four years later, he was actually the first person to write us a check into the company,” Tewari told TechCrunch.

James Jiao, Tewari’s college roommate at Berkeley — now Conversion’s co-founder and CTO — also dreamed of founding his own company, so the two tried building various products, like one for helping marketeers buy product placement ads. They stumbled on the idea for Conversion when they signed up for HubSpot to help them with marketing tasks and decided to build a few extra automation features to layer on top of it.

“It was originally for us,” Tewari said of his startup’s tech. The co-founders enjoyed building their internal marketing tool so much, they wondered if they could sell it and began reaching out to marketing executives for “customer discovery” interviews.

“We actually spent like two months doing like 160 customer interviews with VPs of marketing, 50- to 500-employee businesses, and got a much more positive response than we could have imagined,” Tewari said. 

Techcrunch event

San Francisco
|
October 27-29, 2025

Marketing teams had these tools deeply embedded in their workflows, but everyone had similar complaints about the parts they couldn’t automate.

The duo had found their idea. The family friend introduced them to more marketing execs, which helped them raise a $2 million seed round. At age 19, they dropped out of college to work full time on Conversion.

The founders treated their raised funds so frugally they lived in a two-bedroom, one-bathroom apartment with five other roommates: two people to a room, with people sleeping on the couches and in the closet. 

As they built their product, ChatGPT burst onto the scene. Many legacy marketing automation tools are adding various AI and chat integrations into their wares, but not all of their features support these integrations. Marketing teams wanted “to be able to enrich contacts, [be] able to automate workflows,” for instance, Tewari said. Conversion has baked AI in, which means it can do things like organize leads and automate personalized follow-up emails.

As AI interest has soared, so has the company’s prospects. Conversion is nearing $10 million ARR over the past two years, Tewari said, and about 90% of its customers are midsize businesses that have yanked out a legacy app.

Of course, Conversion is also in a crowded field. Besides the legacy marketing automation tools like HubSpot, Adobe Marketo, or Salesforce Pardot, there are other AI native startups like Jasper, Writer AI, Iterable, Copy.ai, and many others.

But Tewari also has the classic Silicon Valley confidence of a founder in a crowded market. His game plan calls for targeting businesses that use the older marketing tools. Conversion is not, for instance, targeting startups choosing a tool for the first time.

The company has raised a total of $30 million between its seed and Series A, the CEO says, and is doing well enough that the founders have each moved into separate apartments where they have their own rooms, and none of their roommates sleep in a closet.

If you, like practically anyone else with a cell phone in the U.S. and beyond, have received a scam text message about an unpaid toll or undelivered mail item, there’s a good chance you have been targeted by a prolific scamming operation.

The scam isn’t particularly complex, but it has been highly effective. By sending spam text messages that look like genuine notifications for popular services, from postal deliveries to local government programs, unsuspecting victims click a link that loads a phishing page, they enter their credit card details, and that information is swiped and used for fraud.

During a period of seven months in 2024, the scam netted at least 884,000 stolen credit card details, allowing scammers to cash in on their victims’ accounts. Some victims lost thousands of dollars in the scam, researchers say.

But a series of opsec mistakes ultimately led security researchers and investigative journalists to the real-world identity of the maker of the scamming software, Magic Cat, who researchers say goes by the handle Darcula. 

Image Credits:via Mnemonic

As revealed by the Oslo-headquartered security firm Mnemonic and reported in tandem by Norwegian media earlier this year, behind the fluffy cute cat in Darcula’s profile photos is a 24-year-old Chinese national named Yucheng C.

The researchers say Yucheng C. develops Magic Cat for his hundreds of customers, who use the software to launch their own SMS text message scam campaigns at their victims.

Soon after he was unmasked, Darcula went dark and his scam operation has not seen any updates since, leaving his customers in the lurch. But in its wake, a new operation has emerged and is already vastly outpacing its predecessor.

Researchers are now sounding the alarm on the new fraud operation, Magic Mouse, which rose from the ashes of Magic Cat. 

Ahead of sharing new findings at the Def Con security conference in Las Vegas on Friday, Harrison Sand, an offensive security consultant at Mnemonic, told TechCrunch that Magic Mouse has been surging in popularity since the demise of Darcula’s Magic Cat. 

Sand also warned of the operation’s growing ability to steal people’s credit cards on a massive scale.

During their investigation, Mnemonic found photos from inside the operation posted in a Telegram channel that Darcula administered, showing a line-up of credit card payment terminals and videos showing racks with dozens of phones used for automating the sending of messages to victims. 

The scammers use the card details in mobile wallets on phones and conduct payment fraud, laundering their funds into other bank accounts. Some of the phones had mobile wallets overflowing with other people’s stolen cards, ready to be used for mobile transactions. 

Sand told TechCrunch that Magic Mouse is already responsible for the theft of at least 650,000 credit cards a month.

While evidence suggests Magic Mouse is an entirely new operation, coded by new developers and likely unrelated to Darcula, much of Magic Mouse’s success stems from the new operators stealing the phishing kits that made its predecessor’s software so popular. Sand said these kits contain hundreds of phishing sites that Magic Cat used to mimic the legitimate web pages of major tech giants, popular consumer services, and delivery firms, all designed to trick victims into handing over their credit card details.

But despite the prolific nature of Magic Cat and, now, Magic Mouse, and their ability to net millions of dollars in stolen funds from consumers, Sand told TechCrunch in a call that law enforcement is not looking beyond a few scattered reports of fraud or at the wider operation behind the scheme. 

Instead, Sand said, it is the tech companies and financial giants who shoulder much of the responsibility for allowing these scams to exist and thrive, and for not making it more difficult for scammers to use stolen cards. 

As for anyone who receives a suspicious text, ignoring an unwanted message might be the best policy.