EyeTell has been building in stealth for nearly three years. Now it's hiring—and the vision is starting to come into focus.
Chad Hurley doesn't do much press anymore. The man who helped turn shaky home videos into a billion-dollar empire at YouTube has spent the last few years working on something new, and he's been remarkably quiet about it. The venture is called EyeTell, and if you haven't heard of it, you're not alone—the company has maintained an unusually low profile.
What Hurley is building—or trying to build—sits at the intersection of generative AI and short-form video, a space that's gotten crowded fast. Trademark filings, job listings, and scattered corporate records sketch out a company focused on making it radically easier to produce polished video content using AI. There's no product yet, not publicly anyway. But there's money from serious investors, a small team that appears to be expanding, and an ambition that feels both familiar and uncertain.
Whether this becomes Hurley's second act or another Silicon Valley experiment that never quite catches fire is still an open question.
Building Behind Closed Doors
EyeTell incorporated in Delaware and registered in California in June 2023, according to state business filings. As of mid-2025, the company was registered at a WeWork in San Mateo—Hurley as CEO, his brother Brent handling CFO duties and corporate secretary responsibilities. The setup is lean, the footprint minimal.
By late 2023, The Information had reported that the company raised a seed round led by Kevin Hartz's A-Star Capital, with Ron Conway's SV Angel joining in. The dollar figure wasn't disclosed. At the time, sources described the product as generative AI tools for creating short-form video scripts and content from text prompts—think typing out a concept and having the software generate dialogue, scene suggestions, even recordable lines for characters.
The target launch was sometime in 2024. That didn't happen. Now, as winter turns to spring in 2026, EyeTell remains in what insiders would politely call "development mode." The company is hiring aggressively for senior roles, which suggests something is moving. What exactly, and when, is harder to pin down.
The Technical Vision Takes Shape
A trademark application filed with the U.S. Patent and Trademark Office in October 2023 offers one official glimpse of what EyeTell wants to be. The filing covers software for AI-powered video and multimedia content production, plus collaboration tools, development kits, and applications for virtual, augmented, and mixed reality environments. It's a broad claim, perhaps broader than what the company will actually deliver.
Recent job postings get more specific. EyeTell is looking for engineers who can build "real-time mixed media experiences" optimized for Apple hardware—Core ML, the Neural Engine, all the on-device AI infrastructure that Apple has been quietly perfecting. The roles emphasize computer vision, large language models, generative media, and performance that doesn't depend on constant cloud connectivity. There are also openings for a Head of Partnerships, a Head of Support, and a Community Manager. Those aren't the kind of positions you fill unless you're expecting real users soon.
One engineer's public portfolio—visible online before the company presumably asked for it to be scrubbed—described building a platform that orchestrates multiple AI models in tandem: large language models for dialogue, image and video generators for visuals, voice cloning for characters. The output? Real-time comedy skits, generated on the fly. If that's anywhere close to what EyeTell is attempting, the technical lift is substantial.
A Founder With a Mixed Record
Hurley isn't starting from scratch, reputation-wise. YouTube was a phenomenon, and his role in its early days is well-documented. But the ventures since then have been less definitive.
There was MixBit, a video-sharing and editing app launched in 2013 that never gained much traction. Its assets were eventually sold to BlueJeans in 2018. Then came GreenPark Sports, a mobile gaming platform for sports fans that raised $8.5 million in seed funding in 2019 and $31 million more in a subsequent round in 2021. It hasn't become a household name, though it's still operating.
Hurley also holds stakes in the Golden State Warriors and Los Angeles FC, the kind of investments that speak to both wealth and an interest in fan engagement beyond software.
At EyeTell, Hurley has brought on Andrew Gordon as Chief of Storytelling. Gordon comes from Pixar, which is a signal worth noting. Pixar doesn't churn out narrative talent by accident, and hiring someone with that pedigree suggests EyeTell is thinking seriously about how stories get told, not just how they get automated.
The team remains small—probably fewer than ten people—but the hiring push indicates that's about to change.
The Money Behind the Mystery
A-Star Capital led the seed round. The firm was founded by Kevin Hartz, Bennett Siegel, and Gautam Gupta, and closed a first fund north of $200 million in 2021 and was reportedly targeting $300 million for a second vehicle, according to The Information. Ron Conway's SV Angel has backed hundreds of startups over the years, including early bets on Google, Facebook, and Twitter. When Conway writes a check, people notice.
Still, neither EyeTell nor its investors have said much publicly. No press releases, no splashy demos, no website with a product walkthrough. The company has maintained an almost aggressive level of discretion.
There's also a legal matter worth noting. EyeTell is in a trademark opposition proceeding with Bausch & Lomb, the eye care company. The case was suspended as of mid-February 2026, with scheduling milestones extending into the middle of the year. It's unclear whether this is a serious obstacle or just a bureaucratic inconvenience, but it does raise questions about whether the "EyeTell" name will even survive to launch.
A Space That Got Crowded Fast
When EyeTell started raising money, the generative AI video market was still finding its shape. That's no longer true.
Runway, one of the category leaders, raised a $141 million Series C extension in mid-2023 and followed that with a $308 million Series D round in April 2025. It has a live product, paying customers, and a growing reputation among creators. Pika Labs raised $80 million in June 2024 and has continued to expand its user base. Both companies offer text-to-video tools that already work, more or less, for a range of use cases.
Meanwhile, text and script generation has become table stakes. Tools like Jasper, Copy.ai, and OpenAI's ChatGPT are standard equipment for content creators now. The bar for differentiation is high.
EyeTell's apparent focus on Apple's ecosystem—and on-device AI that doesn't require constant cloud calls—could be part of its answer to that challenge. Local processing offers speed, and potentially privacy, advantages. The emphasis on "mixed media" and multimodal AI also suggests the company is trying to do something more integrated than simple text-to-video conversion. Whether that's enough to carve out market share is another matter.
What Happens Next

The hiring activity in early 2026 suggests EyeTell is gearing up for something. Partnerships, support, and community roles don't get filled unless you're preparing to put a product in front of people. But the company has been quiet for nearly three years since it first raised money, and there's no public timeline for when any of this actually ships.
Hurley's track record is, charitably, uneven. YouTube was a once-in-a-generation success. MixBit was a footnote. GreenPark Sports raised real money but hasn't broken out. Whether EyeTell becomes a meaningful player in AI-powered content creation or another well-funded experiment that fades from view is genuinely unclear.
For now, the company is building, hiring, and staying out of the spotlight. In a market moving as fast as generative AI, that's a bet that the right product, when it finally arrives, will be worth the wait. Maybe it will be. Or maybe stealth mode only works if you eventually turn the lights on.
