Singapore robotics company Botsync says it has raised US$11 million in a round led by IvyCap Ventures, taking its total past US$20 million, to build out SyncOS and its MAG mobile robots and enter new markets.
Humyn Labs raises US$47 million Series A for robot training data
EditorialBy TrustList Editorial
Bengaluru company Humyn Labs, which records and labels real-world human activity to train robots and physical AI, announced a US$47 million Series A on 18 November 2026; TechCrunch reports US$26 million equity led by Accel and US$21 million venture debt.
- India
- Robotics
- Machine Learning
- Venture Capital
About Humyn Labs raises US$47 million Series A for robot training data
Humyn Labs raises US$47 million Series A for robot training data
18 November 2026: Humyn Labs, a Bengaluru company that collects real-world data to train robots and other physical AI systems, has announced a US$47 million Series A. According to TechCrunch, the round is made up of US$26 million in equity led by Accel and US$21 million in venture debt led by Lighthouse Canton, with Trifecta Capital. Prosus Ventures, BeeGlobal, Bajaj Finance's Finserv Intelligence and Bajaj Alts AI Opportunities Fund, Dallas Venture Capital, Lighthouse Canton and Jump Trading also joined the equity part.
Not yet independently verified. Humyn Labs' own website announces the US$47 million Series A and gives its head office, founders and team; it carries no release text. The split between equity and debt, the investor names and the revenue and customer figures come from a TechCrunch report of 18 November 2026, and an Indian press report cites a release we could not read. Revenue figures are the founders' own statements to TechCrunch. We will update this when it can be confirmed, and remove this note.
Filming hands at work for robot training
Humyn Labs calls itself "the research and data engine for Physical AI." Robots that work in homes, shops and factories need examples of how people actually handle objects, and video from the internet rarely shows hands, grips and movements in enough detail. TechCrunch reports that Humyn Labs uses its own two-camera recording hardware to film people doing everyday tasks, then labels hand movements and interactions with objects in software. Its website says it works in 36 languages across more than 20 countries; TechCrunch names Brazil, India and South Korea as the main hubs.
The company has its head office in Bangalore, a subsidiary in the United States and a team of about 25 across research, data science, infrastructure, engineering and field operations, according to its site. "Physical AI will only be as good as the real-world data it learns from," says co-founder and chief executive Manish Agarwal on its About page.
Founders and customers
Agarwal previously ran Nazara Technologies, which he took to India's first gaming IPO. His co-founder Ishank Gupta led AB InBev's business in Central and South Asia. The two earlier founded Kratos Studios, now KGeN, and started Humyn Labs in February, TechCrunch reports.
The company told TechCrunch it has 12 paying customers, including frontier AI labs, robotics companies and developers of world models, and that about 90% of its revenue comes from the United States. Agarwal said revenue between April and October was US$10 million, with US$20 million in signed contracts and another US$20 million under negotiation. These are the founders' figures.
What robotics teams can take from it
Training data for physical AI is becoming a market of its own, much as labelled text and images did for earlier AI. Robotics teams buying such data should ask how participants were recruited and paid, what consent covers reuse of the recordings, and whether faces and homes are anonymised, since the data is collected from real people in many countries. Ask also whether the data is licensed exclusively, how it is split between training and evaluation, and in what formats it is delivered, so it fits the team's existing pipelines.
Sources
- Humyn Labs: Announcing $47M Series A, 18 November 2026
- Humyn Labs: About, 18 November 2026
- TechCrunch: Accel backs India's Humyn Labs to supply training data needed by US robotics AI developers, 18 November 2026
Categories & features
- India
- Robotics
- Machine Learning
- Venture Capital
TrustList Weekly
The week in software and IT, in one email
The news that matters to buyers, new rankings and our own research. Every Thursday, free, and easy to leave.
You may be interested in
NVIDIA, Accel, Shastra VC and Micromax join the round. Simplismart sells inference software, not GPUs, and says its annualised revenue run rate grew 14 times in six months.
The San Francisco startup, from Y Combinator's Spring 2026 batch, builds patient-level datasets from scattered hospital records and works in the US, Brazil, India and Argentina.
Inovia Capital led the seed round for the synthetic data company, which launched publicly at NVIDIA GTC in March. Its first app, Studio, is free during a beta open to research teams in waves.
Andreessen Horowitz led the round, with Union Square Ventures and Neo, which led the previous two rounds, taking part. The Watertown company rents out robots that move and pack boxes, billed on use.
a16z and F-Prime lead the Series A, with seed leads Obvious Ventures and E14 Fund following on. Eka claims 1,500 picks an hour, about two and a half times a human baseline.
B Capital and Kindred Ventures lead the round for the San Francisco company founded this year by former leads of NVIDIA's GR00T research; its Dream-1 model drives robot arms costing $2,000 to $10,000.
Its $60 million Series A came in two parts, the first $25 million last November. The company says run-rate revenue passed $100 million in June.
More on TrustList
Everything here links back to the same verified catalogue. Pick your next stop.
- More IndiaThe ranking for this subject
- CompaniesAgencies, consultancies and IT service providers, ranked by verified reviews.
- ProductsSoftware and SaaS with pricing, features, integrations and alternatives.
- AwardsAnnual recognition decided by verified reviews and an independent jury.
- LaunchesNew products and releases, voted up by the community every day.
- AI ModelsBenchmark scores and community ratings for every major model.
- RequestsBuyers describe what they need; vendors respond directly.
- PeopleReviewers, authors and makers with public profiles.
- ComparePut up to four listings side by side before you shortlist.