Pronto Turns Indian Household Chores Into Robot Training Data
Indian home services startup Pronto is using body-camera footage from chores to feed physical AI, while founder Anjali Sardana launches Provision Labs in San Francisco.
SUMO· the number puzzle
Dialo· learn languages by speaking
ediviewer· read EDI files like documents
giftdetect· the right gift in minutesPronto, the Indian on-demand home services startup, is moving into physical AI: footage recorded by its housekeepers during chores is being used to help train robots. Founder Anjali Sardana has also set up Provision Labs, a San Francisco-based venture that will collect data to train and test AI systems built for real-world tasks.
Pronto launched in 2025 as a quick-commerce-style platform competing with Snabbit and Urban Company. It says it has served more than half a million households across 11 cities. In July it formalised its data experiment through Verified, an opt-in premium service in which cleaners and housekeepers wear head- or body-mounted cameras during visits. In October, Provision Labs followed, and days later it announced the acquisition of Mumbai-based Basilisk Industries, a 2023-founded perception technology company. Basilisk founder Shourya Watwani will lead Provision's hardware work and in-house post-processing models. His team is said to have built a 3D hand-tracking model and a first version of a lightweight tactile glove.
Privacy questions remain
How Pronto and Provision Labs work together is unclear. Pronto declined to comment on the point. Earlier reports of the footage collection drew criticism over surveillance of domestic workers. The company says customers are informed several times before recording starts and that personal identifiers are blurred or removed on the device. Raw footage is deleted right after the job, it says, and for 48 hours an anonymised, encrypted recording is visible only to the customer. After that, anonymised footage is used for AI training under its terms. Pronto also says more than half of Verified participants earn over ₹40,000 a month and feel safer with a record of their work, but no independent study backs this up.
Why the data matters
Unlike chatbots, robots cannot learn from text and images online. They need multimodal data on movement, space and object handling. A Goldman Sachs estimate cited by ANI puts the need for general-purpose robots at around 10 million hours of high-quality interaction data.
That has drawn a wave of Indian data-layer startups, including Humyn Labs, EgoData, Egolab.ai, Neocambrian AI, Human Archive and Objectways, which deploy people with cameras, sensor gloves or even full-body suits in homes and factories. It extends India's earlier role in data annotation, now with workers' own movements as the raw material. Questions on pay, informed consent, privacy and ownership of the data remain open as investor interest grows.
Apps worth a look
SUMO: the number puzzleTap the right tiles, add them up, hit the target. Easy to learn, hard to master. Free, for all ages.
Dialo: learn languages by speakingTalk to an AI conversation partner; every sentence comes back corrected and explained in your language. 9 languages, A1 to C2.
ediviewer: read EDI files like documentsOpen EDIFACT and X12 messages, see every segment explained, convert to JSON/XML, export PDF and Excel. Runs locally, no upload.
giftdetect: the right gift in minutesPick who it is for, the occasion and their personality. The search engine reads product descriptions from local shops and ranks what fits. Free, no sign-up.
More from this day
See all →
Calogena wins €40 million EU grant for first commercial SMR
Windows 11 Registry Editor Gets a WinUI 3 Version as a Demo
JRH ASI to Finance Radar System Production for Ukrainian Partner
