Two robotics funding rounds landed on 7 October 2026, and they say something useful about where the money is going. Neither company is selling a humanoid. One sells the data robots learn from; the other sells autonomous freight vehicles that run on existing rail lines.
Mecka AI announced a $60 million Series B led by Sequoia Capital, with Nvidia and Microsoft’s venture fund M12 among the new backers [1][2]. Parallel Systems said it had closed a Series C with $100 million in new capital, led by AVP [3]. Both rounds are confirmed by the companies. Several of the bigger numbers in their announcements are company claims, and we label them as such below.
Mecka AI: what was announced?
Mecka says it has raised a $60 million Series B led by Sequoia Capital, with new investors including Nvidia, M12 (Microsoft’s venture fund), Qualcomm Ventures and Samsung, and continued support from Kindred, Framework Ventures and Neo [2]. It also names angel investors including DoorDash founder and CEO Tony Xu, former ServiceNow and Snowflake CEO Frank Slootman, and Milan Kovac, described as the former head of Optimus at Tesla [2].
TechCrunch confirms the round and its lead investor, and notes that it had previously reported the startup was nearing a new round at a $500 million valuation [1]. Mecka’s own announcement does not state a valuation [2], so that figure should be read as TechCrunch’s earlier reporting, not a confirmed price for this round.
What does Mecka actually do?
In plain terms, Mecka pays people to show robots how to do things.
TechCrunch describes Mecka, founded in 2024, as collecting and analysing human motion data to train humanoid robots and other robots. The company pays people to record themselves doing everyday tasks, such as making coffee or fixing cars, while wearing body sensors and using smartphones [1]. TechCrunch frames the ambition as doing for robotics what Scale AI, Mercor, Surge and other data-labelling firms did for large language models: supplying the human-generated data those systems learn from [1].
Mecka’s own pitch is that the physical world “was never recorded”. Text for language models could be scraped from the internet; the way a hand grips, folds, pours or knows when to let go could not [2]. The company says it has built the whole stack to capture that: its own multi-sensor recording hardware, “capture fleets” recording human demonstrations in homes and commercial settings, and an internal research lab building models for motion tracking, 3D reconstruction and sensor alignment [2]. It says it tested its approach with researchers at Georgia Tech, Stanford, UC San Diego, ETH Zürich, MIT and Meta through a human-to-robot transfer study it calls EgoVerse [2].
The new money, Mecka says, will scale its data infrastructure, expand its research lab and build out commercial robot deployment, where it describes itself as “a modern robotics integrator” that captures data on site and keeps improving a robot after installation [2].
Mecka’s revenue claims: company figures only
Mecka makes some large commercial claims in its announcement. It says it supplies “several of the top frontier robotics labs and multiple Mag 7 companies”, that it passed $100 million in run-rate revenue in June 2026 “within only a few months of operations”, and that it projects a $300 million run rate by the end of the year [2].
These are company statements. Mecka does not name the customers, and run-rate revenue is an annualised snapshot rather than audited annual sales. The year-end figure is a projection. We have not seen independent confirmation of any of these numbers.
Who else is in robot data?
TechCrunch places Mecka in a growing field. It names XDOF as another startup collecting real-world data for robot training, which TechCrunch previously reported was in talks to raise at a $1.2 billion valuation, and notes that human-data platforms that began with language models, such as Scale AI and Micro1, are also moving into robotics [1].
The pattern is clear even without a market-size figure: investors increasingly see the shortage of good real-world training data, not robot hardware, as the bottleneck.
Parallel Systems: what was announced?
Parallel Systems describes itself as a battery-electric autonomous freight rail company. On 7 October it said it had closed its Series C with $100 million in new capital [3]. (The press release headline says “$100M+”; the body text says $100 million.)
New investors include AVP, which led the round, plus Hillspire, Agility Global and Cobalt Capital. Existing investors Anthos Capital, Congruent Ventures, Riot Capital and Collaborative Fund also took part [3]. The company says it has raised more than $200 million to date [3].
Parallel was founded in 2020 by Matt Soule, who the company says spent 13 years at SpaceX [3]. The new funding is earmarked for scaling manufacturing of its third-generation “Panther” rail vehicle, full commercialisation and expansion into international markets [3].
What problem is Parallel trying to solve?
The idea is to make rail competitive for short trips that currently go by lorry. Parallel says its vehicles let railroads handle shorter, lower-density routes, so shippers can use rail for short-haul freight that previously needed trucks [3].
The company frames the opportunity with its own figures: it describes a $1 trillion US surface freight market dominated by trucking, and says routes under 500 miles make up 60% of that market while the rail network is “significantly underutilized” on such routes [3]. Those are Parallel’s numbers, quoted from its release, not independent market data.
How far has Parallel actually got?
This is the part that matters most for an autonomous vehicle company, because regulation is the gate.
Parallel says that since its Series B in 2025 it has moved from prototyping to “live commercial deployment under regulatory test approval” [3]. Its Georgia pilot is running in partnership with short-line rail operator Genesee & Wyoming, under Federal Railroad Administration (FRA) supervision [3]. Chief executive Matt Soule refers to “the FRA-approved pilot proving our platform in real-world corridors” [3].
The company also says “a number of major railroads” are now under contract to deploy its system for commercial freight [3]. It does not name them.
Executive chairman Arun Gupta put it bluntly: “We are now past the question of ‘if’ and fully onto ‘how soon'” [3]. That is a company view, and the release gives no date for moving beyond test approval.
Why put these two rounds together?
Both are bets on the unglamorous layers of robotics. Mecka is betting that robots need far more real-world demonstration data, and that someone will pay to collect it at scale. Parallel is betting that autonomy is most useful where the environment is already constrained, such as a rail line, and where regulators can supervise a pilot step by step.
Neither is a bet on a single impressive demo. Both depend on a slower grind: recording hours of human work, or logging supervised miles on a short-line railroad.
What this does not prove
- Mecka’s valuation. The $500 million figure is TechCrunch’s earlier reporting about the round in progress; Mecka did not state a valuation [1][2].
- Mecka’s revenue and customers. The $100 million run rate, the $300 million projection and the “Mag 7” customers are company claims with no named customers or audited figures [2].
- That human-motion data reliably transfers to robots at scale. Mecka cites a cross-lab study, but its announcement is not a peer-reviewed result, and we have not reviewed the study itself [2].
- That Parallel is cleared for full commercial operation. The company describes its Georgia work as a pilot under FRA supervision and “regulatory test approval”, not full authorisation [3].
- Which major railroads have signed with Parallel, or on what terms. None are named [3].
- Parallel’s market figures. The $1 trillion market and 60% short-haul share are the company’s framing [3].
The Bottom Line
Mecka AI’s $60 million Series B and Parallel Systems’ $100 million Series C are both confirmed by the companies [1][2][3]. Mecka is building a business on paying people to record everyday tasks so robots can learn from them; Parallel is building battery-electric autonomous rail vehicles and running a supervised pilot in Georgia.
The headline rounds are solid. The bigger claims, Mecka’s revenue and customer list and Parallel’s unnamed railroad contracts, are still the companies’ word. The proof to look for next is named customers and, for Parallel, a regulatory step beyond the pilot. None of this is investment advice.
Sources
- Marina Temkin, “Robot data startup Mecka AI nabs $60M from Sequoia,” TechCrunch, 7 October 2026. https://techcrunch.com/2026/10/07/robot-data-startup-mecka-ai-nabs-60m-from-sequoia/
- Mecka, “Mecka raises $60M Series B led by Sequoia Capital” (company announcement; source of investor list, stated use of funds, and revenue claims), 7 October 2026. https://www.mecka.ai/news/series-b
- Parallel Systems, “Parallel Systems Closes $100M+ in New Funding to Fully Commercialize Autonomous Freight Rail System” (company press release via PR Newswire), 7 October 2026. https://www.prnewswire.com/news-releases/parallel-systems-closes-100m-in-new-funding-to-fully-commercialize-autonomous-freight-rail-system-302900326.html
