Blair Silverberg and the problem money couldn’t solve

Before Blair Silverberg wanted to rearrange private finance, he wanted his calculator to do his physics homework. In high school, he learned to program a TI-83 to solve kinematics problems. Years later, with investors wondering whether artificial intelligence would swallow the software industry, he returned to that little machine. It already knew how to add two and two. Why recruit an enormous language model for a job a school calculator could finish between yawns?

The example appears in his July 2026 essay about the continuing value of deterministic code. It is also a useful introduction to Silverberg. He likes a large question, preferably about money, and a concrete mechanism for answering it. He can talk about the allocation of society’s resources, then turn to the machinery that keeps an answer reliable. The ambition travels widely. The arithmetic has to stay put.

At Hum Capital, which he co-founded and previously led as CEO, Silverberg now heads Advanced Analytics. His career has taken him from investing in companies to building a company that evaluates other companies. Along the way, he has had to examine a rather awkward possibility: knowing how investors think does not automatically make them say yes.

A small fortune in stuffed animals

His introduction to markets involved Beanie Babies. His grandmother gave him the toys; his mother helped him sell some on eBay. As he later recalled, unusual tags could turn a stuffed animal into a surprisingly expensive object. The episode had the ingredients of an early financial education: a market, a price, and adults with firm views about how the proceeds could be spent.

His father encouraged investing. By his early teens, Silverberg was buying stocks and reading about markets. The precise childhood chronology matters less than the habit it established. Money became something to investigate. Why did one thing command a price and another fail to find a buyer? What happened after somebody acquired the means to act?

In a later account of his teenage investing, he said he had earned roughly $100,000 in the stock market after starting with bar mitzvah money. That is his account of an early experience, rather than an audited performance record. Its place in his story is clear enough. Access to a modest first pool of capital had changed what he could do.

A child selling collectible toys and an adult designing financial infrastructure occupy rather different markets. Yet both encounter the same practical question: how does a willing seller find a willing buyer? Silverberg’s work would eventually concentrate on the expensive, cumbersome version of that problem, where the merchandise is a company’s future cash flows.

Product design meets the investment committee

At Stanford, Silverberg studied product design. He has described studying under IDEO founder David Kelley, an influence that helps explain his interest in how financing feels to the person seeking it. A capital market can be discussed as an elegant allocation mechanism. To a founder waiting for an answer, it can feel more like a corridor full of closed doors.

His early work included co-founding the tutoring marketplace TutorCloud and serving as a product lead at Intuit. Then came Draper Fisher Jurvetson, where he spent four years as an investor. Software and artificial intelligence were central to that work. Here was an unusually useful seat for somebody interested in how businesses are judged: inside the room where the judgments happened.

Steve Jurvetson became an important connection. The relationship continued after Silverberg left venture investing to build his own business. When Capital, the company later known as Hum, emerged from stealth in 2019, Jurvetson publicly described being eager to work with his former colleague again. One career had supplied both an education and a future backer.

The irony is enjoyable. Silverberg would build a business intended to make financing less dependent on introductions, while benefiting from a consequential professional relationship himself. That does not invalidate the project. It gives the problem a human scale. He had seen what a useful connection could accomplish. The question was how many other businesses deserved consideration without having one.

The former investor gets a no

There is a tempting version of this story in which an investor notices an inefficient market, leaves his firm, and receives a queue of enthusiastic checks. Silverberg’s own account is more interesting. Pitching Hum brought more rejection than he expected. Initially, he treated the resistance as evidence that he had an original idea. Eventually, he began asking what he was missing.

In his writing about that experience, he described communication blind spots, the value of co-founder Csaba Konkoly, and working with a narrative coach. The prospective repairer of fundraising had become a dissatisfied customer of fundraising. Familiarity with the investment committee had provided useful context; it had supplied no immunity to a polite refusal.

His resulting advice gives considerable space to people. Understand an investor’s interests. Give the first conversation room to establish the idea and a later one room to examine its risks. Make the work collaborative. This is a revealing complication in a career associated with quantitative evaluation: Silverberg still cares about whether two people understand each other.

Hum’s founding team also included Chris Olivares, its CTO. In a 2020 conversation together, the two founders explored the software behind evaluating future cash flows. Their shared project joined an investment question to an engineering question. A useful model had to describe a business well enough to inform a real financing decision. An attractive dashboard, by itself, would not pay anybody back.

Giving the numbers a meeting

The company launched publicly as Capital in 2019 and later became Hum Capital. Its underlying proposition was that information already sitting in business software could make financing less laborious. Accounting systems, payment records, and other operating data could help investors understand a business without asking its founders to reconstruct everything by hand.

By November 2021, Hum had expanded its marketplace to support equity fundraising as well as debt. It also announced a second close that brought its Series A to $21 million, with strategic investors including an affiliate of Cowen and Company and Invesco Private Capital. Those developments widened the company’s remit beyond its original emphasis on alternatives to equity financing.

Silverberg compared approaching investors one at a time with booking travel in an earlier era. The useful part of that analogy is comparison. A founder should be able to understand available routes and their costs before choosing one. An introduction can open a conversation. It does not, on its own, reveal whether the proposed financing makes economic sense.

In 2023, Hum introduced SmartRaise, which used connected financial data and investor criteria to provide preliminary debt financing terms. The distinction between a preliminary offer and money in the bank matters. Still, the direction of travel was consistent: bring more of the analytical work forward, so the conversation can begin with something more useful than mutual optimism.

In that year’s interview with Peter Renton, Silverberg described the typical financing range as $1 million to $50 million. These are substantial sums for businesses, yet transactions small enough to make labor-intensive intermediation awkward. His attention settled on the gap between what a company’s records could explain and what the financing process routinely required its people to repeat.

The title gets smaller. The problem stays large.

By October 2025, Silverberg was describing a different working arrangement. Andrew Eisen was running Hum as CEO, and Silverberg had moved into Advanced Analytics. He explained that translating investment experience into code was work he had struggled to find enough time for while leading the whole company. Giving up the chief executive’s chair made room for a particular kind of concentration.

He also admitted that building the business had taken longer than he expected. Two experienced contacts, Wayne Goldstein and Saurabh Khemka, had warned him that a data business might need a decade before things became especially interesting. His younger self had been less receptive to the timetable. Founders are rarely delighted when the answer to their question includes another ten years.

There is a quieter clue to his temperament in the same conversation: he reads annual reports for pleasure. Some people unwind with a detective novel. Silverberg appears content to give the balance sheet a chance. The common attraction may be the possibility that careful attention will reveal something the opening pages did not make obvious.

“read annual reports from investors and companies for fun.”

Blair Silverberg, October 2025

The calculator has the last word

His 2026 writing brings the story back to the TI-83. The essay argues that AI systems will continue to depend on conventional software for tasks with exact rules. Silverberg’s point concerns architecture as much as arithmetic: language models can help interpret intent and call tools, while explicit algorithms handle operations whose answers must be dependable.

That position fits the work he has chosen at Hum. A compelling description of a business still needs contact with its records. A proposed investment still needs its calculations checked. His interest in automation includes the less glamorous question of which parts deserve rules, verification, and repetition. The calculator is a modest object to place in an argument about the future. Modesty is part of its usefulness.

Silverberg’s story now contains a childhood market in toys, a design education, an apprenticeship in venture capital, and the uncomfortable experience of pitching investors who declined. He lives in Austin with his family and works on making private-company information more useful. The early fascination with investing has survived several changes of job title.

What has changed is the scale of the question. He began by deciding what to do with his own small pool of money. Now he works on systems that help other people decide where theirs should go. Somewhere between those two tasks sits the old calculator, offering a pleasingly undramatic standard for financial technology: do the work, get the answer right, and leave the theatrics to somebody else.