Contentize

Premek Hoyetski didn't start as a startup founder—he was a PhD mathematician at Oxford with a clear linear career path ahead: postdoc, professor, tenure. But after spending years in academia, he felt the pull toward something more tangible, faster feedback loops, and real-world impact. Machine learning called to him as the bridge between pure mathematics and entrepreneurship. "I wanted to have more impact on the world, be able to have something tangible, have some faster feedback loop," he explained. The leap meant leaving behind a secure identity and facing years of self-doubt about being a "loser" for abandoning academia.

Premek's first startup, Bring.ai, aimed to disrupt last-mile delivery like Postmates—ambitious, capital-intensive, and ultimately doomed. After a year of bootstrapping with a co-founder, they ran out of cash. His second attempt, Board Technology, was even more avant-garde: quantum computing optimization for logistics. "We were super early to the market," he admitted. The enterprise sales cycle moved glacially, and co-founder disagreements about direction accelerated the company's collapse. Two failures, three years of full-time startup experience, and Premek emerged with clarity: he could execute alone, speed mattered more than big ideas, and he wasn't afraid to fail again.

In early 2019, energized by studying successful founders, Premek created Patacrunch—an automated media outlet where AI journalists scraped data on funded startups, sent cold emails requesting interviews, and published responses with minimal human intervention. It worked. Over 1,000 founder interviews in three months. The platform revealed something profound: media and content could be generated at scale with AI, and audiences would engage.

This epiphany led to Contentize. Rather than editing AI-generated interviews by hand (tedious), Premek asked: why not build a platform where *anyone* could generate articles on demand? In January 2020, with a single Python developer, he built the MVP in two months—a simple interface where users submitted a headline or prompt, clicked "Generate," and received a drafted article. It reached 100 users, proving there was curiosity about AI-generated content.

The first version proved the concept but felt rough. Premek realized the technology was sound; the product experience wasn't. He hired contractors on Upwork and Fiverr to improve the frontend, expand the backend, and integrate additional machine learning models. By June 2020, just three months after the first launch, the second iteration shipped. "From then on we had this great growth of users," Premek recalled. The insight: leverage offshore contractors for specific, well-defined tasks you can't do faster yourself.

The business model evolved organically. Rather than charging purely for SaaS subscriptions (which users resisted), Contentize monetized the content it generated—ads, sponsorships, affiliate links—while offering the platform itself on freemium terms. By September 2020 (nine months in), the company was generating $4,000–$5,000 per month. Most revenue came from the content monetization, a minority from paying platform users—a ratio Premek wanted to flip.

At 25 years old (roughly), running a bootstrapped AI company from Warsaw, Premek embodied the new indie hacker: leveraging AI breakthroughs (cheap APIs from big tech, open-source models), hiring globally (no expensive San Francisco salaries), and iterating faster than anyone with a traditional team could. He was talking to investors about seed funding, but on his own terms. When asked if he'd found a co-founder, he demurred: "I'm fine solving the whole thing because it's more interesting from my perspective... if there's any lack of skills, I'm happy to hire someone and maybe give them a small portion of equity, but not on a co-founder level." Contentize was generating 100,000 articles per month. The platform had become proof that the future of media wasn't about hiring writers—it was about architecting systems where AI, algorithms, and human judgment worked in concert.