Hardware – Edge Machine Learning Engineer

Mecka announces $60M Series B for robotics AI data. Last Updated: 8 October 2026 Co-founder of JustAINews Key Points * Mecka develops robotics training data from human movements and helps businesses deploy robots in real commercial environments. * Mecka raised $60 million led by Sequoia, with NVIDIA, M12, Qualcomm Ventures, Samsung, Kindred, Framework Ventures, and Neo. * Mecka reported a $100 million annualized revenue run rate in June 2026 and targets $300 million by year-end. Credits: Sequoia Capital What Is Mecka? Mecka is a robotics AI company founded in 2024 that develops data infrastructure and deployment solutions for physical AI. It records human movements and interactions using specialized sensors, then converts that information into datasets that help robots learn real-world tasks. Its services include data collection, AI model evaluation, robot integration, and commercial deployment. Mecka serves robotics research laboratories, AI model developers, and enterprises adopting robotic systems. The company operates data collection programs worldwide and offers EgoVerse, a platform for accessing and analyzing human activity datasets. Mecka's founding team includes Josh Gao, Mogen Cheng, Jason Chong, and Duy Nguyen. Gao and Cheng previously worked together on a financial technology startup serving restaurants, while Chong founded a cryptocurrency exchange that was later acquired by Coinbase. Mecka's $60 Million Series B Funding Mecka raised $60 million in Series B funding led by Sequoia Capital, with new investors including NVIDIA, M12 (Microsoft's Venture Fund), Qualcomm Ventures, and Samsung. Existing backers Kindred, Framework Ventures, and Neo also participated. The company reported exceeding a $100 million annualized revenue run rate in June 2026, within a few months of beginning operations. It expects that figure to reach $300 million by the end of the year. A revenue run rate estimates annual revenue based on a shorter period of business activity, rather than revenue already earned over a full year. Why Robots Need Data From Human Movements AI systems learn by studying examples. Language models can use large collections of written material, while robots need information about movements, physical contact, pressure, and the position of objects. Much of that information needs to be recorded directly from the physical world. Consider a robot learning to fold clothes. It needs to understand where to place its hands, how firmly to hold the fabric, and how to adjust its movements as the material changes position. These details are difficult to capture through ordinary text or video alone. Handling objects accurately remains a challenge in robotics. Humanoid robots still face difficulties with precise movements, flexibility, and independent decisions. Researchers are developing datasets that capture touch and movement information to help machines interact with objects more effectively. These systems have potential uses in manufacturing, warehouses, hospitals, and homes. Better training methods could help businesses introduce robots into workplaces where tasks vary and machines must respond to changing conditions. How Mecka Plans to Use Its $60M Funding Part of the investment will support additional recording equipment. Mecka designs and manufactures its own devices, which contain multiple sensors that capture different aspects of human activity. Its recording teams operate in homes and commercial environments, collecting examples of people performing physical tasks. Research is another priority. Mecka develops computer vision systems, which help computers interpret visual information, and AI models that combine information from several sensors. The resulting data includes hand movements, three-dimensional positions, and measurements that help describe how people interact with objects. According to the company, its technology can track hand positions with accuracy better than one centimeter in real-world conditions. This information is used to prepare datasets for robotics researchers and developers. Mecka also plans to expand its work with commercial customers. It provides robotic equipment, collects data at customer sites, adapts AI models to specific tasks, and handles installation and ongoing operations. The company continues collecting data after robots are deployed, allowing their models to be trained further using information from the workplace. How Mecka's Robot Training Technology Works Mecka's technology covers three main activities. First, its sensors record human movements. AI software then processes those recordings, identifying details such as hand positions and how objects move in three-dimensional space. Finally, quality control systems prepare the information for use in robot training. The aim is to turn everyday human actions into data that robots can learn from. For example, researchers can use recordings of someone handling an object to train a robot to perform a similar task. Mecka has also worked with researchers from Georgia Tech, Stanford, UC San Diego, ETH Zürich, MIT, and Meta through EgoVerse, a study investigating how robots can learn from human demonstrations. The experiments involved different laboratories, robotic systems, and tasks. The company says it supplies several robotics research laboratories and multiple companies belonging to the Magnificent Seven group of major technology businesses. Investors Behind Mecka's Series B The funding also attracted several individual investors, including Tony Xu, founder and CEO of DoorDash; Frank Slootman, former CEO of ServiceNow and Snowflake; and Milan Kovac, who previously led Tesla's Optimus humanoid robotics program. Mecka had secured financing earlier in the year. In June 2026, the company announced a separate $60 million funding round led by Framework Ventures, with participation from Menlo Ventures, SV Angel, and Kindred Ventures. The latest financing will allow Mecka to produce more training data, continue its research, and increase the number of robots deployed in commercial environments. Funding details. * Company name: Mecka * Funding round: Series B * Date: October 2026 * Funding amount: $60 million * Lead investors: Sequoia Capital Get the industry's biggest AI news straight to your inbox.

Mecka AI reveals $60-million USD Series B round backed by Sequoia, Nvidia. Startup aims to provide the real-world training data required to deploy robots. Toronto- and New York City-based Mecka AI has raised a $60-million USD ($86-million CAD) Series B round led by Silicon Valley's Sequoia Capital. Mecka announced the financing on Wednesday. It was supported by new backers like chip giant Nvidia, the corporate venture capital arms of Microsoft and Qualcomm, Samsung, and existing investors Kindred Ventures, Framework Ventures, and Neo. DoorDash co-founder and CEO Tony Xu, ex-ServiceNow and Snowflake CEO Frank Slootman, and the former head of Tesla Optimus Milan Kovac also participated. Mecka aims to build "the data and deployment layer" for physical AI by providing real-world data to train robots. The two-year-old startup also claimed it surpassed a $100-million USD run-rate in June and expects to hit $300 million by the end of this year. Mecka aims to build "the data and deployment layer" for physical AI, providing real-world data sourced from body sensors and iPhones to train humanoids and other types of robots. Demand for this type of data - which, unlike the text, images, and video required to train large language models, cannot be scraped from the internet - is booming. Mecka said this capital will help the company scale its data infrastructure, deepen its internal research lab, and build out its commercial robot deployment layer. TechCrunch first reported last month that Mecka was raising a new Sequoia-led round at a valuation of about $500 million. The new funding comes months after Mecka acquired small, bootstrapped Vancouver-based startup Docula and announced $60 million USD in Series A financing. Mecka secured that amount across an initial, $25-million USD Series A round from last November and a $35-million extension that closed over the summer, financed by Framework, Menlo Ventures, SV Angel, Kindred, and others. Mecka is betting that video captured from a first-person human perspective will help robots learn faster and scale more efficiently. The company collects this data from home settings, culinary work, chemistry labs, task platforms, metal fabrication, and leather shops, and converts it into training-ready data for multiple frontier robotics labs and Big Tech firms. The startup has built hardware replete with sensors to capture this data and the computer vision and multimodal models to turn it into usable training fodder for robots. While Mecka is domiciled in the US, three of its four co-founders are Canadian, and the vast majority of its staff are in Toronto. AI claimed nearly half of all venture dollars invested in Canada last year, and global interest in robotics, humanoid makers, and firms building embodied AI is heating up. The Canadian Robotics Council has clocked "a rapid increase" in the number of robotics startups being launched, both at home and abroad. Vancouver's A&K Robotics and NorBot, Winnipeg's TetraGen Robotics, and Montréal-based Mecademic are among the other startups that have benefitted from this surge. UPDATE (10/07/26): This story has been updated to include additional information shared by Mecka AI. Feature image courtesy Mecka.

Mecka has raised a $60 million Series B led by Sequoia Capital to build an internet-scale dataset for training robots. The New York and Toronto startup also added NVIDIA, Microsoft's M12, Qualcomm Ventures and Samsung as backers. The round brings Mecka's total disclosed funding to $128 million across three rounds since August 2025. The company reported an annual run-rate revenue exceeding $100 million as of June 2026 and expects to reach $300 million by year end. Mecka collects egocentric data by recording human motion from a first-person perspective using body sensors, smartphones and specialised hardware. Participants perform demonstrations of everyday activities in real-world settings. The resulting dataset, called EgoVerse, targets developers working on humanoid robots and embodied AI. The startup employs between 40 and 60 people.