The Mecka AI Sequoia valuation of roughly $500 million is still being finalised, but the fundraising trajectory of this two-year-old startup tells a story that goes well beyond one pending term sheet. Three months after closing a $60 million Series A, the New York-based motion-data company is in advanced talks with Sequoia Capital for a new round that would more than triple its last disclosed post-money mark.
According to Dealroom, Mecka AI’s June 2026 follow-on closed at a $140 million post-money valuation. The Sequoia-led deal, if it closes near the reported figure, would represent roughly a 3.6x step-up in a matter of months. Dealroom’s figures are estimates, and the terms of the current round are not yet final.
From Sneakers to Sensors: The Founders Behind Mecka AI
Mecka AI was founded in 2024 by four entrepreneurs, none of whom came from a robotics background. Josh Gao and Mogen Cheng, both Canadian, previously ran a restaurant fintech startup. Jason Chong sold a crypto exchange to Coinbase. The fourth co-founder, Duy Nguyen, handles operations and, according to Gao, built his fortune flipping shoes as a self-described sneakerhead before joining the venture.
What the team lacked in robotics credentials, they compensated for in pattern recognition. They identified a structural problem: the physical-world data required to train general-purpose humanoid robots was scarce and poorly organised. Their answer was to build the data layer from scratch.
The company pays people to record themselves performing everyday tasks (making coffee, fixing cars, loading shelves) using body sensors and smartphones. The approach, sometimes called egocentric data capture, mirrors what Scale AI and similar human-data platforms did for large language models: industrialise the collection of the raw material the models need.
Mecka’s name is a riff on “mecha,” the fictional genre of giant robots piloted by humans. The reference is apt: the startup’s entire proposition is that human motion, recorded at scale, is the missing input for machines that need to move through the physical world.
A Capital Stack Built Fast, With Mecka AI Sequoia Talks at the Top
The company’s fundraising history has moved quickly. An $8 million seed round, led by Neo with participation from SV Angel, A-Star Capital, and others, gave it the runway to prove the model. Home robotics firm 1x Technologies signed on as an early partner, and the startup claims paid contracts with several Fortune 100 corporations.
The $60 million Series A, as reported by Fortune, was actually assembled in two tranches: a $25 million round that closed in November 2025 and a $35 million follow-on completed earlier in 2026. Framework Ventures led the round; Menlo Ventures, SV Angel, and Kindred Ventures participated, as did Ted Xiao, a former Google DeepMind researcher and founding member of Jeff Bezos’s AI venture Project Prometheus.
Gao told Fortune at the time of the Series A announcement that the company was projecting a $100 million annual run rate by the end of 2026. He was specific about the basis: those projections were anchored to already-signed contracts, not pipeline estimates. He declined to name the customers.
Framework Ventures, for its part, reinforced its commitment to the sector. The firm closed a $400 million fourth fund in June 2026, explicitly expanding its focus into robotics and AI, according to the firm’s Framework Ventures news page.
Mecka currently employs between 40 and 60 people, spread across New York and Toronto, according to a summary of its current operations. The company also has a subsidiary recorded by PitchBook in Vancouver, Canada.
A Sector Where Valuations Are Running Ahead of Revenue
Mecka is not alone in attracting capital at speed. Rival data startup XDOF, which collects real-world physical data through a different method, is in talks for a Series B led by 8VC at a $1.2 billion valuation, according to TechCrunch reporting. XDOF’s annualised revenue was approaching $50 million at the time those talks were reported.
The comparison illustrates the dynamic now shaping the robot training data sector: investors are pricing in the eventual scale of demand from humanoid robotics programmes, not just current revenue. Mecka’s valuation, if the Sequoia round closes, would be five times its annualised revenue target, assuming it hits the $100 million run rate on schedule.
The deal terms are still subject to change, and Sequoia declined to comment. The next data point will be the closing announcement itself, and the customer names that may or may not come with it.
