Rashid Khasanov typed the first seven hundred rows of his database by hand, in debt, straight after the last company took everything he had

Everybody says start a database business. Almost nobody types the first seven hundred rows.

Rashid Khasanov did, one investor at a time, sitting in New York with his savings gone and his credit cards maxed out. The company before this one was an SEC-registered robo-advisor, which is to say he picked, as a first venture and as a person who does not write code, a regulated financial product with a compliance burden that defeats funded teams. It took everything. In his own words: "I lost almost everything, drained all my savings, maxed out credit cards, and went deep into debt."

The thing he could not do during all of that was find investors. Not because they did not exist, but because the information about who they were and how to reach them was scattered across a hundred places and stale in most of them. So the next thing he built was the list he had needed. That is the whole origin: the failure did not teach him to aim lower, it told him exactly which spreadsheet everybody was keeping by hand.

Four products, no code, $42k a month

Angel Match now holds more than 125,000 investors, with verified email addresses on over 99,000 of them, and does $37,300 a month. Around it sit three more: Investor Hunt at $4,557, Pur Social at $491, and Journalist Hunt, a hundred thousand journalists in a searchable list, at $180. Four businesses, about $42,000 a month, bootstrapped, run by somebody who cannot write the software.

The strategy is stated plainly and it is better than most funded companies manage: "curated databases are harder to kill." While the entire industry ships wrappers that a model update can obsolete in an afternoon, he is compounding an asset that gets more valuable and more expensive to copy every month somebody maintains it. And maintenance is the actual job. A hundred and twenty-five thousand rows decay constantly; people change firms, funds close, addresses bounce. Keeping 99,000 verified emails verified is unglamorous, endless, and precisely the moat.

His filter for what to build next is one line long: find the thing people are already doing by hand in a spreadsheet. Then he does his own customer support, because that is where the next product is hiding.

The part he says out loud

"Healthy and rational thinking people would have quit a long time ago." He is joking, and he is also describing the mechanism. Between the robo-advisor collapsing and the portfolio clearing forty thousand a month, there were years of small tools, iterations, and things that did not work, funded by nothing, while the debt from the last attempt sat there.

It tracks with a kid who made kites, built bicycles and bred parrots to earn money. That is somebody who was never going to be talked out of running his own thing, and the robo-advisor is the same instinct pointed at the hardest target in the room. Both of those are the same person, and the second one is only possible because of the first.

The goal now is $100,000 a month, mostly by getting churn down. Whether you find databases boring is not the point. Four of them, profitable, built without an engineer, by a man who was in debt six years ago, is one of the more genuinely instructive things happening in software right now.

Curated databases are harder to kill.

Rashid, on why he builds the least fashionable kind of software there is