Mecka AI, based in Toronto and New York City, supplies training data for real-world robot deployment and has completed its second acquisition by purchasing Ottawa-based Thirdspace. The mobile app, described as a mix between BeReal, Instagram, and Snapchat, was built to encourage spontaneous in-person connections. The deal closed a few weeks ago alongside Mecka's Series B funding and brings Thirdspace's cofounders into leadership roles at the startup.
Acquihire Strengthens Mobile Growth Team
Mecka special projects lead Mark Grinev said Thirdspace's small but mighty team was the most important part of the transaction. The cash-and-stock deal, for which both firms declined to disclose a price, is effectively an acquihire similar to Mecka's earlier purchase of Vancouver-based Docula. Thirdspace cofounders Cedric Osagie and Eshan Betrabet have joined Mecka as heads of mobile capture, though the app will remain operational.
Thirdspace's Rapid Rise
Launched last year by former Carleton University students, Thirdspace lets users post photos of activities and invite friends to join. Images are only briefly viewable, and accepted invites reveal a location for capturing a photo together. The app has surpassed 20,000 downloads, 60 million views, and tripled monthly active users with about 100,000 Canadian dollars in angel funding.
Scaling Egocentric Video for Robotics
Mecka is betting that first-person video will help robots learn faster and scale more efficiently. Unlike text, images, and video used to train large language models, this egocentric data cannot be scraped from the internet. The approximately 80-person startup has built hardware with sensors and multimodal models to convert video into training material for frontier robotics labs and Big Tech firms.
A Pipeline Built on Consumer Hardware
Mecka's approach relies on individuals using smartphones and wearable sensors to record first-person footage of daily activities. The startup pays contributors and transforms raw video into structured training data with computer vision. This pipeline is designed to supply the specialized datasets that frontier robotics labs and major technology firms increasingly demand.
A 100 Million Hour Target
Mecka has already collected millions of hours of egocentric video from hundreds of thousands of individuals. According to Grinev, that remains far from enough to train real-world robots, so the company has set a target of 100 million hours by the end of 2027. Reaching that goal will require millions of people to download its app, strap phones to their heads, and submit footage of everyday tasks.
Applying Consumer Growth Tactics
The founders of Thirdspace will help Mecka apply consumer growth strategies to its data collection app. Osagie said Mecka is the first robotics team he has seen growth-hacking its app like a traditional consumer product, which gives it incredible advantages. Grinev added that the pair are exceptional growth hackers who have captured lightning in a bottle for going viral.
Funding and Expansion Plans
The acquisition was announced shortly after Mecka closed US$60 million in Series B funding, equivalent to 86 million Canadian dollars, from Sequoia Capital, Nvidia, Microsoft, Qualcomm, and Samsung. Grinev declined to share the exact valuation, despite reports that the round valued the company at about US$500 million. Mecka has now raised US$120 million and plans to use the capital to grow teams in Toronto, New York, and San Francisco while covering compute, storage, and data capture payouts.
Talent Wars and Future Acquisitions
Grinev acknowledged that the AI talent wars have become quite intense. While Mecka does not have any other acquihires planned at the moment, he said he would not be surprised to see the company pursue more such deals in the future. The startup claimed a run-rate above US$100 million in June and expects to reach US$300 million by the end of this year.
Mecka's acquisition of Thirdspace shows how robotics data startups are borrowing growth tactics from consumer apps. By bringing in founders with viral growth experience, Mecka hopes to close the gap between its current video library and its 2027 target. The deal, combined with fresh funding, positions the company to compete aggressively in the emerging market for robot training data.
Source: betakit