When Kai Zhao pitched investors on VideoTutor last fall, he was barely old enough to drink. But the age gap didn't seem to trouble YZi Labs, which led an $11 million seed round for the AI education platform in what the firm called its first software bet.
Perhaps more surprising than the founder's youth is where his company's lessons are landing: TikTok. VideoTutor, which generates personalized instructional videos from student questions, had racked up over 50 million cumulative views on the platform as of May 2026, according to company figures. That's an unusual distribution strategy for edtech infrastructure—a category that typically courts school districts and corporate training budgets, not scroll-happy teens.
Whether those views translate into sustainable revenue remains an open question. But the Sunnyvale startup's approach reflects a broader tension in education technology right now: how to build something students actually want to use, in a market where venture funding has contracted sharply and consumer attention spans have never been shorter.
The Backers and the Bet
YZi Labs announced the October 2025 round alongside Baidu Ventures and Amino Capital, with investor blogs also listing Jinqiu Fund—a ByteDance-affiliated investor—and BridgeOne Capital as participants. The deal sheet reads like a who's who of Chinese and Silicon Valley tech money, all wagering that AI-generated instruction can scale where human tutors can't.
Zhao had already logged time at a Y Combinator-backed AI education startup before launching VideoTutor under the corporate name Imaginix Inc. His co-founder, James Zhan, comes from stints at Meta and Google. Data aggregator SignalBase reported an earlier $1 million pre-seed round in May 2025, though the company never formally announced that figure and it remains unconfirmed.
How It Works
The mechanics are straightforward enough. Students submit a prompt, upload a screenshot of a problem, or type out a question. VideoTutor's backend—LLM orchestration layered with Manim-based programmatic rendering—spits out an animated, voice-guided explainer. Think Khan Academy, but generated on demand and tailored to the specific confusion of a single calculus student at 11 p.m.
The company says students can interrupt mid-lesson to ask follow-ups or request different explanations. It's a pitch that emphasizes "concept-level understanding" over simply serving up answers, though skeptics might note that's also what every edtech platform claims before pivoting to homework help.
In its first ten days live, according to a November 2025 blog post from YZi Labs, VideoTutor logged around 20,000 users and churned out more than 20,000 videos—early metrics that may have grown substantially since. The startup also fielded over 1,000 API integration requests from organizations curious about embedding the technology—though how many of those inquiries converted into actual partnerships is unclear.
The TikTok Play
Fifty million views is a big number. It's also a company-reported, unverified number, which is worth flagging. Still, it signals intent: VideoTutor isn't just building for institutions. It's chasing virality.
That consumer-first tilt sets it apart from competitors focused on selling to schools or tutoring centers. But it also raises questions about business model. TikTok fame doesn't always pay the bills, particularly in education, where monetizing free content has bedeviled startups for years.
VideoTutor says it's received "purchase and deployment inquiries" from Tencent Group and the Xiaotiancai smart-watch ecosystem—company-reported claims that neither party has confirmed. The company's LinkedIn profile lists somewhere between two and ten employees—nine were visible as of mid-April—though that count may have shifted since.
Context and Competition
The timing is tricky. Overall edtech venture funding slumped through much of 2025, according to HolonIQ data, even as AI-specific tools continued to attract outsize interest. In January, language tutoring platform Preply raised €127 million, a reminder that investors remain willing to write checks for AI-enhanced learning models—provided the unit economics make sense.
VideoTutor's ambitions are broad: smart-learning hardware integrations, content platform partnerships, institutional tutoring systems. Whether a small team can execute on all of that while managing TikTok growth and enterprise sales simultaneously is another matter.
For now, the company appears to be threading a needle. Build something students share. Figure out how to charge for it later. It's a familiar playbook in consumer tech, though one that's littered with cautionary tales—particularly in education, where the gap between engagement and revenue can be wider than founders expect.
