Kamua

Paul Robert Cary had already built Findie, a Netflix-for-short-films platform, during his MBA at Columbia and London Business School. While the platform initially aimed to be a discovery destination for curated independent films, early success led brands like L'Oréal, Nissan, and Red Bull to request white-label versions. This pivot to B2B was profitable but revealed a critical pain point: videos needed to be converted from widescreen to vertical formats for mobile apps. Paul realized his team's custom AI-powered automation tech—originally built to generate hundreds of thousands of video trailers—could be repurposed to solve this universal problem.

Paul elevated Radu Amarie, his CTO from Findie, to co-founder status to lead product development. Rather than just automating the conversion (which would be 50-99% accurate), the team made a deliberate product decision: they invested heavily in building an intuitive manual editing interface for the "last mile." Paul explains the philosophy: customers receive an automated boat ride to their island but must swim the final distance. The key was making that swim easy. The interface takes 20 minutes to learn, compared to 50-200 hours for traditional desktop software like Adobe Premiere Pro. Everything was built for the web browser—no downloads, no local computing required—which solved a real customer problem: their first customer from Omnicom had slow, old laptops for social media managers who needed to move fast.

Their first paying customer came directly from their previous business—Omnicom needed exactly what Kamua solved. Paul then proactively reached out to Google and shared the product vision. Despite Google working on similar technology (AutoFlip, released open source), they welcomed Kamua into their Google Cloud for Startups program and provided $100,000 in cloud credits. This led to partnerships with Nvidia (Inception program) and HubSpot (Startups program), giving Kamua infrastructure cost advantages and credibility. They deliberately chose to restrict access to paying "Supporters" rather than open it to everyone—inspired by Superhuman's exclusive model—to filter out curious tire-kickers and retain only serious video creators.

The most effective growth channel has been content marketing: posting tutorials on YouTube and Reddit that explicitly help people solve problems with competitor software (Adobe). Rather than just teaching Kamua, they position themselves as solving creative people's broader problems, building brand awareness among their target audience. They've gained hundreds of new social followers monthly this way. What didn't work: an open alpha period attracted curious users with no real video editing needs, distracting the team from their core mission. Pricing is kept simple: entry-level at $21/month for annual subscriptions, with higher tiers. Their first B2B customer pays $700/year per user on an annual prepaid contract. One customer reduced a 16-hour video editing process to 15 minutes—a 64x speedup that compounds value when they can now create 3 videos instead of 1 for A/B testing.

At +$6,000 MRR with a six-person team, Kamua is lean and focused. Paul invested $225,000 of his own money, a decision he credits with maintaining focus—he had to sell his vision to get talent and investor support rather than coast on personal wealth. The biggest challenge remains separating signal from noise: paying customers request unnecessary features, investors suggest pivots outside their expertise, and competitors' hype campaigns can be distracting. Their next major feature will revolutionize editing: edit the transcript, and the video automatically removes those deleted sections. For entrepreneurs, Paul's advice is to practice persistent thoughtfulness—slow growth is fine, pivoting with conviction is fine, but giving up before finding product-market-fit is not.