Mecka announced a $60M Series B in October 2026 led by Sequoia Capital, with new investors including NVIDIA, Microsoft's M12, Qualcomm Ventures and Samsung. Existing investors Kindred, Framework Ventures and Neo also participated, alongside angels including DoorDash CEO Tony Xu and former ServiceNow and Snowflake CEO Frank Slootman.

TechCrunch reported in September that Mecka was negotiating a financing at a $500M valuation. The October announcement confirmed the financing but did not disclose a valuation, so the reported $500M should not be treated as a confirmed post-money figure. Fortune had separately reported a $60M Series A in June 2026.

Mecka builds the data and deployment layer for physical AI. It designs multi-sensor capture hardware, operates collection fleets in homes and commercial environments, and processes the recordings into motion tracking, 3D reconstructions and synchronized sensor data. Its customers use these signals to train robots on physical tasks that text and ordinary internet video cannot describe completely.

Its Egoverse dataset captures first-person human activity across tasks and environments. Customers can browse and visualize data, query datasets, use inference APIs and commission task-specific collection. Mecka also offers a mobile application for collecting demonstrations, complementing its own sensor hardware.

The company is expanding from supplying training data into integrating robots in commercial settings. It combines third-party hardware and models with on-site capture, post-training, integration and ongoing operation. This gives enterprises a way to deploy physical AI without assembling an internal robotics team, while producing new demonstration data from the deployment environment.

Mecka sells dataset access and commissioned data collection to robotics labs, frontier model developers and enterprises. Its website describes entry points ranging from access to existing data through task-specific collection, model evaluation and full data programs. Public sources do not disclose its pricing or the recurring share of its contracts.

The economics depend on converting a distributed collection operation into usable training signals. Contributors and capture fleets generate the raw recordings; Mecka's hardware and research infrastructure determine whether those recordings have the alignment, motion detail and quality that a customer's model requires. Reusing a capture system and processing pipeline across customers could improve efficiency, while specialized task requests add collection and quality-control work.

Commercial integration broadens the offering beyond datasets. An enterprise deployment can require hardware selection, environment-specific training, installation and ongoing operations. These services could deepen customer relationships and generate proprietary on-site data, but they also make delivery more operationally demanding than licensing a fixed dataset.

Mecka competes with other suppliers of robot-training data, including XDOF, and with human-data companies expanding beyond language models into robotics. Scale AI and Micro1 can bring existing customer relationships and data-production operations to the market, while robotics-focused suppliers can specialize in physical capture and human-to-robot transfer.

Mecka's differentiation is its control over capture hardware, collection operations and reconstruction models. It can design the sensors around the signals a model needs and deliver structured motion data rather than rely only on video. Its Egoverse work with academic and industry researchers provides a route to demonstrate whether human recordings transfer usefully to different robots and tasks.

Robotics companies can also build collection infrastructure internally. The decision to outsource depends on the required task coverage, the quality of supplied data, and the time and cost of reproducing that capability. As Mecka moves into deployment, it also competes with systems integrators and internal enterprise robotics teams.

Mecka's initial market is training data for robotics and embodied AI. Physical models need demonstrations across objects, environments and edge cases, creating demand for task-specific collection and evaluation as labs move from research demonstrations toward commercial deployments.

The deployment business extends Mecka into the spending required to put robots to work. Supplying capture, post-training, integration and operations could let the company participate in a broader customer budget than data licensing alone. Each deployed environment can also become a source of additional training and evaluation data.

The expansion depends on robots becoming reliable and economical enough for enterprises to adopt. Mecka's role as an integrator lets it work across hardware and model vendors, but its growth still depends on those systems producing useful work at a cost customers will pay.

Data transferability: Human demonstrations do not automatically produce reliable robot behavior because robot bodies, sensors and operating environments differ. If customers cannot turn Mecka's data into measurable task improvements, demand for additional datasets could weaken.

Operational costs: Designing hardware, coordinating capture fleets and reconstructing recordings require sustained investment and quality control. Expanding into robot deployment adds installation and ongoing service obligations that could limit margins as revenue grows.

Customer dependence: Large AI and robotics labs can develop their own collection systems or work with competing suppliers. Mecka has not disclosed customer-level revenue concentration, making it difficult to assess how much its run rate depends on individual buyers or projects.