They went from nearly bankrupt — savings gone, $15,000 in credit card debt, co-founder Daniel Dhawan sleeping on a mattress in a friend’s apartment — to $100,000 in revenue in five days. That turnaround seeded a $2.8 million round led by Andreessen Horowitz’s new Speedrun program, with Hustle Fund’s Elizabeth Yin, ChapterOne, Founders Inc., Austen Allred, Expo’s Charlie Cheever and Evan Bacon, and Runway’s Siqi Chen joining fast.
It all started with a single tweet that changed everything. Rork is a mobile app builder that lets anyone with limited technical background create fully functional native iOS apps from nothing more than simple text prompts. On February 12, after months of work and one pivot, co-founders Levon Kvirkvelia and Dhawan launched Rork with a single tweet.
“We were completely at a disadvantage. We ran out of money quickly,” Dhawan told TechCrunch. They received a small check from angel investor Matt Shumer, co-founder and CEO of OthersideAI (maker of AI writing tool HyperWrite).
Their first tweet got some traction, but the breakthrough came February 24, when people on X were discussing a competing product from a company called Bolt. Shumer posted that Rork was better.
I was blown away. Rork lets you create full iOS apps just by describing them! Zero code. This changes everything about app development. Rork is better than Bolt (yes, I invested immediately after trying it).
Shumer attached a video of Rork, and the post went viral — over 1 million views. Usage spiked. But the founders were in financial distress: they’d each put $15,000 on credit cards to keep the AI running.
“We were basically investing our own money in AI because we hadn’t raised much,” Dhawan said. “We were almost out. Then fifteen minutes after that post, Austen Allred invested $100,000.” On Shumer’s viral day, Founders Inc. and Hustle Fund’s Yin committed, plus referrals from other angels. “So day one we raised $350,000. It was insane.”
The Pivot That Saved the Company
It sounds like instant success — and in some ways it was. But for the founders it felt like rescue from the brink. This was each founder’s third self-started company; they built hit mobile apps as teenagers. Now 25 (Kvirkvelia) and 27 (Dhawan), they’d already spent most of the cash earned from Rork’s predecessor — a Cursor-like vibe coder for non-technical users.
Dhawan had been in San Francisco since December attending Founders Inc. events and fundraising. He worked through Christmas and slept on a YC-founder friend’s apartment mattress. Kvirkvelia stayed home in Georgia building the product. Fundraising was going nowhere — every investor meeting ended with polite passes and no term sheets, while the runway measured in weeks rather than months.
Then competitor Lovable launched and went viral. Rork wasn’t ready to compete. “Our prototype was done, and then they launched,” Dhawan recalled. “We were really disappointed.” So Kvirkvelia convinced Dhawan to pivot: instead of another AI web-coding tool like Lovable, return to their roots and build the Lovable for mobile apps.
Nobody had done this well before — because while web development is hard, building native mobile apps is “ten times more complex,” Kvirkvelia told TechCrunch. “We’re the Lovable for Expo, and the Lovable for React Native,” referencing the two dominant mobile frameworks. When competitor Bolt later launched a mobile vibe-coding product, the founders felt déjà vu and chose to launch the same day as Bolt.
After Shumer’s viral tweet and the angel wave, a new investor introduced the founders to Andrew Chen at Andreessen Horowitz, general partner of a16z’s new Speedrun program — a 12-week mentorship for early-stage startups with $5 million in partner credits (AWS, Google Cloud, OpenAI, Microsoft, NVIDIA, Stripe, Deel) and typically up to $1 million in investment. Chen reached out, but Dhawan didn’t immediately commit — he had a pre-seed term sheet from another firm. Chen, determined not to lose the deal, ran a fast interview and submitted a competing offer. The founders accepted and will join the July 28 program.
Daniel and Levon are technical powerhouses with deep understanding of mobile development and distribution, which lets them move fast and build a great platform. They’re exactly the kind of founders a16z Speedrun wants to back.
Better than the funding: paying users are arriving. Two months after Shumer’s viral tweet, the two-person team’s ARR hit $550,000. Dhawan has moved out of his friend’s apartment and into his own place.
Why the Story Matters Beyond the Outcome
Behind the viral-moment narrative is a structural insight about the AI coding market. Web vibe-coding is now a crowded category (Lovable, Bolt, v0, Replit), but mobile vibe-coding is technically harder — native iOS requires Xcode, Apple’s review process, device testing, and framework-specific patterns that web tools don’t handle. The 10x complexity claim is credible: anyone who has shipped a React Native app versus a Next.js site knows the difference. By pivoting to mobile before the web-coding gold rush saturated, Rork found the less contested territory — and Shumer’s tweet provided the distribution that technical differentiation alone couldn’t. The deeper lesson for founders in the AI-coding space: the category’s technical ceiling (native mobile > web) determines where the open territory is, and being ten times harder is sometimes a better moat than being first — and the ARR trajectory from zero to $550,000 in sixty days with a two-person team is the kind of evidence that makes that argument for you.
