How Vishal Virani built Rocket from zero to half a million users in 16 weeks

Executive overview

Most funded founders come with Stanford or IIT pedigree. Vishal Virani had neither — he grew up in Surat with no engineering role models and no money. He built two companies anyway, absorbed 40+ VC rejections, and pivoted from a Figma-to-code tool into a full AI software platform the moment GPT-3 landed.

Build the right product first; the funding, users, and traction follow.

Background and mindset

• Grew up in Surat — largest diamond manufacturing hub, nine of every ten diamonds processed there
• First engineer in his family; parents gave blessings but had no money to support him
• Applied farming-world lesson: you cannot control rain or weather, only your response to it
• Read everything on the internet rather than lamenting the absence of an Ivy League network
• Location and zip code are irrelevant — use available resources instead of building excuses

First venture: DeWise (Figma to code)

• Founded post-graduation; chose building over a master's degree, committed two years to learning over earning
• Identified that developers were drowning in repetitive work (LinkedIn flooded with hiring posts)
• Built a Figma-to-code converter: users upload designs, receive production-ready code
• Scaled from zero to 350k users; processed 10 million Figma screens — highest in the category
• GPT-3 launch changed everything — 60 hours spent in a single session exploring what was possible
• Conviction: Figma-to-code would become a minor feature; the real opportunity was full software development automation

Pivoting to Rocket

• Pivoted from DeWise to Rocket: user inputs a prompt, platform delivers a complete, production-ready application
• Differentiated from competitors (new web-coding platforms launching daily on Product Hunt): rivals produced non-scalable prototypes; Rocket targeted production-grade output
• Early signal: a solo non-technical founder emailed to say he closed seven projects in 20 days — started at $25/month, grew to $4,000/month with zero sales contact
• Reached half a million users and millions of applications on the platform within 16 weeks of launch

Fundraising: 40+ rejections to $15 million

• 40+ consecutive VC rejections when every other idea seemed to be getting funded
• Pattern: investors rewarded Stanford, IIT, ex-Google, ex-Microsoft — Virani had none of those signals
• With 15 days of cash left, received first term sheet
• Response to rejection: analysed pitch mechanics daily — word choice, numbers, articulation — rather than sinking into frustration
• Salesforce Ventures discovered Rocket via an X post, tried the product, and reached out unprompted
• Secured $15 million seed in 15 weeks of inception — one of India's largest early rounds at the time
• Raised from Excel and Salesforce Ventures; no formal pitch required once the product spoke for itself

Product and growth principles

• Early-stage allocation: 60–80% effort on product, 20% on marketing — not the reverse
• Build real product, not pitch decks; non-native English speakers especially should prove rather than persuade
• Cursor vs. GitHub Copilot: Microsoft's full backing didn't prevent Cursor from capturing the market with a better product
• Email every user personally (even automated); read replies; provide direct support — trust compounds fast
• Filter signal from noise: hundreds of pieces of advice circulate on X daily — identify which signals apply to your specific context
• Every pain point is diagnostic — ask what mistake you made that caused the pain, then fix it