Nikil Viswanathan has said that Alchemy was actually his team's sixteenth pivot. By his own telling, they had spent about nine years on startups with little to show before 2021.
One of them nearly made it. Down To Lunch, a social app he co-founded around 2016, began as a Sunday afternoon project. The first version was built in a single night, with deliberate technical debt so it could ship at once. It solved a personal itch: meeting friends spontaneously. Social media reviews spread it. Then, in under 48 hours, the app reportedly lost 90 percent of its growth. The team also dressed up as Berkeley students to get feedback and recoded around what they heard.
Ethereum had shown what Web3 could be, but building on it was painful, Greylock says.
He grew up in a small Texas town few people have heard of, he has said in a Stanford talk. Other profiles say Lubbock and Chicago for birth, which should be treated cautiously. He studied computer science at Stanford, reportedly for both degrees, and there met Joe Lau, with whom he taught an online database course. Alchemy's own company page says Lau holds Stanford bachelor's and master's degrees and worked at Pinterest, Google and LinkedIn.
By 2017 the two had a new angle. Ethereum had shown what Web3 could be, but building on it was painful, Greylock says. A developer had to run servers just to read the chain. Alchemy offered to run them. The pitch became the AWS for blockchains.
What almost nobody saw coming was the speed. After years of near-misses, the 2021 boom sent builders to Alchemy. $80 million at a $505 million valuation in April 2021, $250 million at $3.5 billion in October, then $200 million at $10.2 billion in February 2022. A Stanford summary of his 2023 talk says it went from public launch to a $10 billion valuation in 16 months. Adobe and OpenSea were customers.
The cost was a market cycle. The sources read contain no valuation news after February 2022, so $10.2 billion is a high-water mark, now about 33 months old. In 2026 the company hired an AWS veteran, Bill Platt, as COO, and tear-sheet data put headcount at 384.
The sources are thin on the personal side: no interview read here describes a family reaction, debt or a night of doubt. What they do record is a count. Sixteen versions, one that took off and then slid in two days, and a partner from Stanford who stayed. For a first-time founder, the number to notice is not ten billion. It is sixteen, and the fortune arrived to people who were still there.