Atmosfy launches Atlas to license its place-tagged video to AI companies

Why it matters

Atlas takes Atmosfy from a consumer discovery app toward licensing real-world video as AI infrastructure. Its commercial case depends on proving that its large, company-reported corpus is current, accurately mapped and cleared for customer use.

Atmosfy launched Atlas on October 7th, a platform that packages location-tagged videos of real places as data for AI, search and commerce products. Founder and CEO Michael Ebel is extending the video-discovery business he built for people choosing where to go into a licensing business for companies building systems that make those choices for them.

Ebel's starting point was personal: he had worked as a bartender and saw how harsh online reviews could affect local businesses. A Berkeley Haas profile says that, while working at Meta, he saw video as a fuller way to show what a venue was like than static ratings and text reviews. During the pandemic, Atmosfy's early pitch was that customers could post videos showing local businesses were open and worth visiting. Ebel had also founded LookFwd, an earlier app for coordinating outings and finding local offers, according to a Haas account.

Atlas moves that idea from consumer discovery toward infrastructure. Atmosfy says its video map contains more than 500 million experiences across 100 million places, 20,000 cities and 150 countries. Those are company figures; the announcement provides no independent audit or breakdown of how many videos are recent, license-ready or available for a particular commercial use.

Atmosfy describes Spectra as its intelligence model. The company says Spectra turns videos into time capsules that connect location and moment, creating an updated picture of places. Its Atlas materials describe the corpus as geo-tagged, structured and licensed, and present potential uses including model training, retrieval and recommendation. The announcement does not specify Atlas's technical interface, pricing, customer deployments or which frontier AI developers have signed licensing agreements.

Atmosfy named existing relationships with Capital One, Expedia, SeatGeek, GetYourGuide, Virgin Hotels, Resy and Funko, but did not describe which of those organizations use Atlas, what they license or whether they are customers of the new product. Its website illustrates Atlas in a travel-discovery setting, but an illustration is not evidence of a commercial deployment.

The path to monetization also depends on what rights Atmosfy can offer. The company's terms of service say users grant Atmosfy rights to sublicense submissions for commercial uses, including AI development and model training. They also say contributors are not entitled to additional compensation for that licensing unless a separate written agreement provides for it. The terms establish the rights Atmosfy publishes for user submissions; they do not reveal how many videos in the claimed corpus are covered or the scope of any customer licenses. For an AI data business built on contributor video, rights provenance is part of the product.

Ebel's bet builds on a company that has already raised outside capital. Atmosfy announced a $12 million seed round in August 2023, led by Redpoint Ventures, with participation from Palm Tree Crew Investments, Streamlined Ventures, Industry Ventures, Canaan Partners, Village Global, Progression Fund and Convivialite Ventures, according to its funding announcement. A contemporaneous Haas magazine account described the round as $14 million, leaving a public discrepancy between the two figures. The Atlas announcement names additional backers, including Capital One Ventures, DNX Ventures and Suntory Ventures, but gives no total raised, valuation or new financing tied to the launch.

Ebel's experience spans more than consumer apps. Public profiles describe prior roles at Meta and PwC and service as a U.S. Air Force captain; Haas's account connects his time at Meta to the video-first thesis behind Atmosfy. With Atlas, Atmosfy is selling to businesses and AI developers, which must decide whether its footage is current, accurately matched to places and cleared for their intended uses.

AI assistants that recommend a restaurant, hotel or event need visual information about what a place looks and feels like, alongside its name and address. Atmosfy is positioning its archive as a visual layer for those recommendations. The announcement names no AI-model customer, licensing deal terms or Atlas revenue, so it establishes the direction of the business without demonstrating commercial traction.