Gleb Budman and the price of remembering

Gleb Budman had helped start a backup company when his own laptop failed. It was February 5, 2008. The service had not yet reached its friends-and-family beta, but it could already do the thing that mattered: bring a file back. Budman became the first person in Backblaze’s early company timeline to restore an otherwise unrecoverable file. The chief executive had acquired a customer’s education ahead of schedule.

For most people, backup is a chore with an invisible reward. Nothing happens, and that is the point. Budman and his colleagues chose to build a business around this awkward proposition: pay us regularly so that a future calamity becomes an inconvenience. Their original unlimited plan cost five dollars a month. Behind that modest price sat a problem large enough to occupy the next two decades.

Today, Budman is Backblaze’s co-founder, chief executive and board chair. The company trades on Nasdaq. In June 2026, it announced a five-year, $335 million storage agreement with CoreWeave. It is a considerable distance from a broken laptop, although the distance becomes easier to understand when you follow the bills.

The engineer before the cloud

Budman was born in Saint Petersburg, Russia, and moved to the United States at five. His family’s journey took them through Austria and Italy to Pennsylvania, followed by a move to Silicon Valley after his father found an opportunity there. Budman grew up speaking English, with Russian spoken by his parents. He later described himself as an American citizen.

His first technical interests were physical. He took things apart to understand them; reassembly had a less dependable success rate. Cars, boats and robots held his attention. At Berkeley he studied mechanical engineering, and his first job after college was at GE in the nuclear industry. For a time, he considered a robotics business that would automate manual work in that industry.

The internet redirected him. He earned an MBA at Berkeley’s Haas School of Business in 1999 and accumulated experience at Kendara, Excite@Home, MailFrontier and SonicWall. These were useful places to learn that a business could change owners, direction and prospects with disconcerting speed. Kendara was acquired by Excite@Home; MailFrontier went to SonicWall.

When he later recalled the Excite period, the memorable detail was organizational confusion. The people who had bought the company left, and the newcomers were asked what they wanted to do together. An acquisition could apparently arrive before anyone had settled the agenda. By the time Backblaze began, Budman had seen enough of the conventional startup route to have opinions about taking it again.

An apartment with a storage problem

The spark came from a friend of co-founder Brian Wilson whose laptop had died. Wilson asked about her backup. She wanted her data recovered, rather than advice about the precautions she should have taken. Further conversations suggested the problem extended well beyond one unlucky computer owner.

Budman and four co-founders committed to working for a year without salaries. They began in Wilson’s one-bedroom Palo Alto apartment. The arrangement was meant to be temporary. More colleagues arrived. Eventually Wilson moved out, and the team acquired the use of his bedroom. Silicon Valley has many accounts of companies expanding into new markets; this one first expanded into a sleeping area.

The group initially considered different ways to solve backup, including better software for external drives. Once they pursued an online service, they needed somewhere to put the files. Buying storage from existing providers would make the proposed economics fail. An affordable subscription would have to support the actual cost of holding customers’ data.

So the team built infrastructure. The red Backblaze Storage Pod became a recognizable piece of the company’s working life: a dense arrangement of commercially available drives in a custom enclosure. The eventual hardware hid a succession of experiments. Tim Nufire later recalled an early system of USB drives that lasted roughly an hour under production load. A cloud could have a surprisingly short first day.

Budman’s place in this story is alongside the people doing that engineering. Wilson, Nufire, Casey Jones and Billy Ng were fellow founders with distinct contributions. The company’s origin depends on that group, their previous working relationships and their willingness to keep solving an expensive problem while spending little.

A radio button pays the bills

One of Budman’s most revealing business stories concerns a signup form. Backblaze already sold monthly and annual subscriptions. The monthly option was selected by default. Switching the default to annual billing changed what customers chose, and therefore when cash arrived.

In his 2013 account, 60 percent of new customers stayed with the annual default. For every hundred new subscribers, the revised mix brought in $3,200 up front, a 40 percent improvement in cash collected. The engineering effort was negligible. A tiny interface decision had changed the flow of money into the company.

“All it took was a five-minute change.”

Gleb Budman, on the billing default

He also explained the catch. More cash up front did not mean more profit. An annual payment came with a year’s obligation to provide service. Servers, drives, power, bandwidth and support would still need paying for after the first payment had been spent. The trick worked only if the commitments and the costs could be made to fit.

That qualification matters. Budman was describing the mechanics of staying in business, including the less photogenic portion where somebody has to ensure the cash lasts. It gives his public writing a useful texture. A reader can admire the outcome and still inspect the arithmetic.

His approach to investors had similar conditions. In July 2012, Backblaze announced $5 million from TMT Investments after five years of bootstrapping. Budman explained that the team had reconsidered outside funding every six months. They wanted room to hire ahead of cash flow and grow faster, while preserving a culture built around efficiency. Taking substantially more money raised a question for him: would abundance change the habits that had made the business work?

Giving away the instructions

Backblaze’s thrift produced something unexpectedly public. In 2009, the company released its Storage Pod hardware design, including the information other people needed to build their own. A piece of infrastructure developed to keep costs down became a way for strangers to understand the company.

The original design article drew more than a million readers, according to a 2015 retrospective. The logic was practical as well as generous. Backblaze wanted attention, and it wanted economical storage. Nufire later said the team would have welcomed someone else building the equipment better and cheaper. Selling boxes had never been the central ambition.

Readers and customers began asking for a different kind of access: could they use the storage platform directly? On September 22, 2015, Budman announced B2 Cloud Storage. Developers and other users could store, retrieve and share data through a service built on the infrastructure behind backup. Years spent solving an internal problem had created another business.

Going public on November 11, 2021, supplied a more ceremonial milestone. Budman had become board chair that January. Yet his subsequent account of the IPO returned to familiar ground: what had this complicated exercise cost, how did it work, and why was there so little practical guidance for the next founder attempting it?

In February 2023, he introduced Open Sourcing the IPO. Backblaze had spent more than $10 million in connection with the listing. He proposed sharing what the team had learned about advisers, staffing, documents and the chief executive’s responsibilities. Even the bewilderment could be reused.

There was a related change that July. Shareholders voted to eliminate the company’s dual-class share structure, removing the unequal voting powers attached to its two classes of stock. Budman explained the decision in terms of fairness, accountability and the company’s readiness to move beyond its early public-company arrangements. The promise of openness had reached the voting rules.

The same question, many more zeros

Budman’s personal websites offer a lighter counterpart to the boardroom material. One calls itself a placeholder on the internet. An older site preserves albums of hang gliding, skiing and Burning Man. His fifth-anniversary company history includes the discovery that pigeon rental could become an expense for an April Fool’s video. The storage business evidently left a little room for the absurd.

The current business asks for a different scale of attention. Backblaze’s June 2026 CoreWeave announcement describes multi-exabyte capacity supporting portions of CoreWeave’s managed storage infrastructure, including hard-drive-based storage tiers. The $335 million agreement runs for five years. It is a commitment to supply infrastructure over time, with the work of delivery still attached.

For Budman, the connection to AI runs through data that needs somewhere economical to live between and around demanding computing tasks. Training, checkpoints, model outputs and other workflows create storage requirements. The company’s second-quarter 2026 results reported B2 revenue growing 34 percent year over year, while overall revenue grew 18 percent.

The figures are larger now, and the customer conversations have changed. The thread back to the apartment is the persistent question of what reliable storage should cost. Budman once experienced the answer as a recovered file on a failed laptop. His company’s next chapter depends on making that same practical confidence available to customers with considerably more to remember.