Back to directory
AI & ML · Physical AI

Mecka AI

The data and deployment layer for physical AI

carbon based, ceo @meckaai
3.2K followers
TLVC Rating
Hook
Editing / Creativity
Copy
Sentiment of launch
Distribution strategy
Community Rating
No ratings yet
Your rating
Sign in to rate this launch.

About

Mecka AI is building a data and deployment layer for physical AI, supplying robotics teams with the human movement datasets and infrastructure they need to train and ship real-world systems. The company collects data from humans using custom hardware like body sensors as well as iPhones , then turns that capture into labeled, model-ready material for humanoid and general-purpose robotics developers. The launch matters because the field has shifted from demos to deployment, and teams now need a partner that handles capture devices, annotation, and integration in one stack rather than stitching vendors together. The company was co-founded by CEO Josh Gao, CTO Jason Chong, Duy Nguyen, and Mogen Cheng , and it has grown quickly since coming out of stealth. Mecka AI is headquartered in New York City and has raised a total of $60 million across two rounds , with Framework Ventures leading and Menlo Ventures, SV Angel, and Kindred Ventures among the investors who participated . Early customers include home robotics company 1X Technologies as an early partner , alongside contracts the team has signaled with larger enterprises. This round is aimed at scaling the data engine described in the launch video, including multi-sensor capture devices, an in-house computer vision and machine learning effort focused on scene reconstruction and simulation, and the operational footprint needed to deploy robots into varied environments. Gao has framed the company beyond data alone, with his startup now employing 40 people and aiming to help firms integrate and train robotics models directly rather than acting as a pure data aggregator . For founders and operators watching the embodied AI stack, Mecka is positioning itself as the picks-and-shovels layer underneath the robot companies themselves.
Tags
<500KExplainerSeries AB2BGlobalUSVertical AIFunding announcementFounder-led
Comments (7)
Sign in to join the discussion.
Priya Raghavan1d ago

Called the data layer for physical AI back when it was a deck with three slides. Watching this thesis crystallize is doing something to me.

Samiul Karim1d ago

got a perception lead coming off a self-driving gig, 6 years sim-to-real, looking specifically at physical AI infra plays. who do I send them to over there?

Tunde Adebayo1d ago

ok wait, deployment layer AND data layer? pick a lane or build the whole stack, but the tweet is trying to do both in one breath.

Clara Voss1d ago

Free rewrite: 'Mecka: ship robots, not spreadsheets.' You can Venmo me.

kenji1d ago

the launch video had actual robots doing actual things instead of stock footage of a server rack, rare W for the genre.

Adaeze N.1d ago

Quick one: where does the training data from customer deployments live, and who owns the derived models when a robot leaves a factory floor? Asking for every GC about to sign your MSA.

Misha Sokolov1d ago

any chance the SDK ends up public, or is this strictly enterprise-call-us-for-pricing energy? curious how a hobbyist would even touch this.