Editorial illustration for Pronto Founder Enters Physical AI With Provision Labs; Data Use Unclear
Pronto Founder Launches Physical AI Data Startup
Pronto Founder Enters Physical AI With Provision Labs; Data Use Unclear
Anjali Sardana is running two companies now. The founder and chief executive of Pronto, the Indian household services startup, has launched Provision Labs, a San Francisco venture building data infrastructure for AI systems that handle physical tasks in the real world. Sardana will keep her job at Pronto while taking this one on too.
Provision Labs has backing from Lachy Groom, General Catalyst, Bain Capital Ventures and Glade Brook Capital, according to the company's website. It hasn't said how much it raised, what it's worth, or who its first customers are. The Economic Times reported, citing people familiar with the venture, that Provision Labs plans to collect data to train and test AI systems operating robots and machines outside controlled lab settings, working with developers across robotics and other fields.
The timing raises an obvious question. Pronto already runs Pronto Verified, a program that records household service visits with wearable cameras so the footage can be anonymised and fed into robotics training. Whether Provision Labs taps into that same pipeline, or builds something separate, hasn't been addressed by either company.
According to The Economic Times, Provision Labs will collect data to train and test AI systems that use robots and other machines to perform tasks in real-world environments. The report, which cited people familiar with the venture, said the startup will work with developers building models for robotics and other applications.
Why this matters
Sardana running Pronto and Provision Labs at once raises the obvious question: whose data is this, actually. Pronto's household-service recordings and Provision Labs' stated plan to "collect data to train and test" physical AI systems look like they could be the same pipeline wearing two names, but nothing on either company's site confirms or denies it. For developers building on physical-task datasets, that ambiguity matters more than the funding list.
Backing from Lachy Groom, General Catalyst, Bain Capital Ventures and Glade Brook Capital signals investors believe in the data-infrastructure thesis for robotics and embodied AI, but belief isn't disclosure. We'd want to know, before any of this data ships into a training set, whether Pronto's household workers and customers consented to having their sessions used for a separate for-profit AI venture. The Economic Times' line about collecting data "to train and test" is thin on specifics, and Provision Labs hasn't said how it sources data independent of Pronto.
Until that's clarified, treat this as two companies sharing a founder and possibly a lot more, not as a case study in clean AI infrastructure.
Related Reading
Business & Startups Nordic pilot adds Gemini for Education, NotebookLM to boost AI literacy
Business & Startups Kling launches Video O1, all-in-one model with MVL bridge using transformer
Business & Startups DeepSeek Seeks More Capital Weeks After USD 7B Funding Round
From Archives Anthropic launches free OSS Scanner; maintainers report inflated severity
From Archives Anthropic: Single AI Agent With Skills Outperforms Multi-Agent Systems
Common Questions Answered
What is Provision Labs and what does it focus on?
Provision Labs is a San Francisco-based venture founded by Anjali Sardana that builds data infrastructure for AI systems handling physical tasks in the real world. The company collects data to train and test AI systems that use robots and other machines to perform tasks in real-world environments, working with developers building models for robotics and other applications.
Who are the investors backing Provision Labs?
Provision Labs has secured backing from several prominent venture capital firms including Lachy Groom, General Catalyst, Bain Capital Ventures, and Glade Brook Capital. These investors are supporting the company's mission to develop data infrastructure for physical AI systems.
Why is there ambiguity about data ownership between Pronto and Provision Labs?
Anjali Sardana is simultaneously running both Pronto, an Indian household services startup, and Provision Labs, which creates uncertainty about whose data is being used. Pronto's household-service recordings and Provision Labs' stated plan to collect data for training physical AI systems could potentially be the same pipeline, but neither company's website confirms or denies this connection, leaving developers uncertain about data sourcing.
What is the significance of Sardana holding dual leadership roles?
Sardana's simultaneous leadership of both Pronto and Provision Labs raises questions about potential data sharing between the two companies and whether Pronto's existing household service data could be leveraged for Provision Labs' physical AI training. This dual role creates ambiguity that matters significantly to developers building on physical-task datasets, as they need clarity on data origins and usage.
Further Reading
- Maxinsights Raises $30 Million Seed Round for Physical AI Data - Startup Researcher
- Pronto founder Anjali Sardana sets up US-based physical AI venture Provision Labs - The Economic Times
- Pronto founder Anjali Sardana sets up US-based physical AI venture Provision Labs - The Economic Times
- Pronto founder Anjali Sardana sets up US-based physical AI venture Provision Labs - WN.com
