Before he was Allan Grant, he was Mikhail Sapov, a boy in Mykolaiv with a programmer's patience and an entrepreneur's borrowed vocabulary. One day, at five, he kicked a ball into a neighbor's flowers. His grandmother scolded him with a phrase that sounded like an insult: “You little capitalist.” His father offered a different definition. Capitalism, he explained, meant that two people could create things of value, trade them, and both benefit. The boy's response was immediate: that sounded great.
The exchange stayed with him because business in the Soviet Union was more than a career choice in the Sapov family. It was a forbidden instinct. His father longed to build companies and, after Ukraine became independent, stood in line on the first day private businesses could be registered. He called his company Allan Grant. When the family moved to the United States in 1994, father and son both adopted the name. For 10-year-old Mikhail, it was a symbol of an American future and a promise to keep the entrepreneurial project going.
Code supplied the other half of the identity. Grant has said he wrote his first game at seven. From 13 to 18, he spent thousands of hours building online games in C for communities with thousands of players. Long before growth dashboards and referral funnels, he was learning the basic bargain of software: make a system, release it to people, watch what they do.
The expensive first draft
At Georgia Tech, Grant started Webmasters International, an outsourcing firm with web developers in Atlanta and Mykolaiv. By graduation, it employed 50 people, 40 of them in Ukraine. The company was profitable and real. It was also fragile in ways a young founder did not yet understand. When it unraveled, Grant was left with roughly $500,000 in debt. Paying it back took six years.
One scene captures the speed of the reversal. Grant had reached the finals of a business competition with a professor's recommendation. A month earlier, the company had appeared healthy. By the day of his speech, it was in trouble. He fired 25 people from his hotel room, walked before the judges, told them what had happened, and said he did not deserve to be there.
Debt closed off the tidy escape route. Grant concluded that an ordinary salary would not clear the liability, so he kept starting companies. That could sound like bravado from a distance. Up close, it was arithmetic. Failure did not give him a motivational slogan. It gave him a long repayment schedule and a sharper sense of what evidence looked like.
Three months in a house
Talkable began in 2010 with a constraint that now reads like a reality show premise. Grant met two co-founders through a founder-matching event. They rented a house, moved in together for three months, and tried to discover what they should build. The promising clue was Dropbox. Its referral program rewarded both the person sharing and the friend joining. Grant and his partners wondered whether online stores would pay for the same mechanism.
They built a version and sold it to a store on the first day. Within a week, more shops wanted it. The company, then called Curebit, entered Y Combinator's Winter 2011 batch with sales already moving. Its early product treated referral marketing as an optimization problem. A merchant could vary the reward, wording, shared message, landing page, and design. The system would test the combinations and learn which ones converted.
An early finding was wonderfully inconvenient for marketers who preferred the obvious lever. The size of the discount did not always matter as much as the words around it. A referral offer was a small social negotiation. The software worked by testing the psychology, not simply increasing the coupon.
The line is not tidy. The through-line is repeated contact with customers, candidates, and the people writing the software.
A recruiting tool escapes the building
The second major company arrived as an answer to Talkable's own hiring problem. Strong engineers were receiving so many messages that cold outreach barely registered. Grant and his collaborators imagined a private marketplace where employers would disclose salary offers before an interview. Other companies could see the competition and bid for the same candidate.
They did not begin by polishing the marketplace. In one month, they recruited 88 developers. Grant then took the idea to investors connected with Y Combinator and 500 Startups. Within a week, 120 companies had joined. Only when both sides were waiting did he sit down for a 36-hour coding marathon. He finished the first site, launched it, and left for Burning Man.
The project was called DeveloperAuction before it became Hired. Grant served as co-founder and CTO and led product until December 2016. The launch sequence remains the useful artifact: assemble a market, then build the machinery. The code sprint gets the cinematic detail, but the quieter work had already removed the largest risk. Developers and employers had raised their hands before the site was complete.
Find a buyer before refining the product. Talkable made a sale on its first day.
Recruit both sides of a marketplace before investing in a finished platform.
Test language and presentation. Price is only one variable in a human decision.
Put product judgment where the work happens, with context and authority attached.
The office that owned the vision
Talkable's path to Ukraine included a public mistake. In 2012, while testing a homepage, the company temporarily copied the design of Highrise, the 37signals product. David Heinemeier Hansson called it out. The resulting criticism followed the young company into recruiting, where some Rails developers declined offers because of its reputation.
Grant searched through open-source communities and found Bogdan Gusiev, then a prominent Ruby on Rails contributor in Kyiv. The two began talking. Gusiev eventually built a team in Ukraine, and the San Francisco engineering plan faded. Grant was sensitive to a common failure of distributed companies: one office thinks while another receives tickets. He framed the arrangement differently. “The vision for our product springs directly from the Kyiv office,” he said.
The philosophy showed up in hiring. Grant described five filters: ambition and the desire to improve, quality of work, cultural fit, references, and conduct during the interview. He preferred small independent teams and competitive compensation to headcount collected for display. This was the scar tissue of the first company made operational. Growth had to survive contact with payroll, trust, and the quality of the people doing the work.
Home, interrupted
Grant's connection to Ukraine was personal before it was organizational. He first returned at 18, initially wondering why he should visit a country his family had worked so hard to leave. The feeling changed when he arrived. The people and atmosphere felt close. He began spending more time there and, in 2021, moved to Kyiv. “Deep down I'm still Misha,” he said in an earlier interview. “I consider myself a Ukrainian.”
When Russia launched its full-scale invasion in February 2022, Talkable had about 60 Ukrainian employees. The company had prepared salary advances before the attack. Afterward, it converted those advances into bonuses, accelerated later payments, protected jobs, helped coordinate relocation and lodging, and joined a direct-aid campaign that raised more than $40,000 by that April. It also sought displaced Ukrainian candidates for open roles.
Grant remembered how the tone of calls changed. During the first two weeks, every meeting began with people asking how each other was doing. By the end of the third, colleagues pushed back: they were present, alive, and ready to talk about the agenda. Work offered a piece of ordinary time, provided it remained a choice rather than a demand.
There is no clean founder lesson to extract from war. There is only the record of prior decisions becoming consequential. A team given strategic ownership before a crisis remained a team during it. The remote structure, the pay practices, and the years of shared context became practical support when abstractions failed.
The builder returns
After stepping away from Hired's daily product work in 2016, Grant spent a year traveling through more than 30 countries. He made time for music and classical literature. The pause did not hold. Industry conferences and meetups pulled him back, including an event in Lviv where he found two Talkable executives the company had sought for more than a year. By 2018 he was again involved in strategy, business development, and product. Today, Talkable lists him as founder, CEO, and chairman.
His public biographies use the words hacker, entrepreneur, and adventurer. The sequence matters. Grant's recurring move is to make something small enough to test: a game, a referral offer, a talent marketplace. The results have included a collapsed company, a long debt, a product born in a shared house, and a second startup launched before a desert trip. First principles, in his version, are less philosophical than procedural. Put the idea in front of someone. Watch closely. Keep what survives.