In 2012, Radha Ramaswami Basu noticed something that most AI investors were too busy building neural networks to see: someone had to label all the data those networks needed to learn from. The unglamorous, meticulous, never-photographed work of telling a machine "this is a tumor," "that is a stop sign," "here is where the lane ends." She had spent 34 years in technology by that point and she knew, with absolute certainty, that the field was sitting on a labor vacuum.
Her answer was iMerit Technology - a company that has, with quiet precision, become load-bearing infrastructure for the AI programs that capture headlines. Autonomous vehicles, medical imaging, generative AI, financial fraud detection. If a foundation model was trained in the last decade, there is a reasonable chance that somewhere in its training pipeline, iMerit's annotators touched the data.
What Radha built at iMerit is stranger and more interesting than a standard Silicon Valley success story. The founding workforce did not come from Stanford or MIT. It came from the Anudip Foundation - a nonprofit she co-founded with her husband Dipak in 2005 to train youth from low-income households in India in English language, workplace readiness, and technology skills. When iMerit needed people who could do precise, patient, detailed annotation work, Radha realized something: the young women Anudip had trained in embroidery had the exact fine motor precision and attention to detail that computer vision annotation demanded. Precision is precision, whatever medium it occupies.
That insight - social enterprise talent meeting frontier AI demand - is the through-line of Radha's career. She is not a technologist who discovered social impact. She is a social impact builder who recognized a technology opportunity hiding inside one of the world's most overlooked talent pools.
Before iMerit there was Support.com, which she led as CEO from 1999, navigating the company through its NASDAQ IPO in 2000 and a secondary offering in 2003 - doing it through the dot-com crash, which requires a different kind of nerve than doing it in a bull market. And before Support.com there was Hewlett-Packard, where she spent two full decades. In 1985 - a time when multinational tech companies had barely begun imagining what India could become as a software hub - Radha flew to Bangalore and built HP's first India software center. She grew it into a $1.2 billion operation. This was not incremental execution. It was institution-building from scratch in a city that was still figuring out what an IT industry could look like.
She grew up in Tamil Nadu, attended an all-girls school run by Irish Catholic nuns, and then enrolled at the College of Engineering in Guindy - an all-boys institution at the time. Her father believed in mathematics for his daughter, which in 1970s India was itself a small act of defiance. From there she completed a Master's in Computer Science and Biomedical Engineering at the University of Southern California before joining HP in 1978.
Today iMerit's 7,400-person workforce spans data annotation, model evaluation, red teaming, content moderation, and what the company calls "expert annotation" - work done by domain specialists rather than general-purpose crowd workers. In 2025, Radha launched the Scholars program: an invite-only global network of PhDs, MDs, lawyers, linguists, and engineers brought in to annotate the kinds of datasets that require genuine expertise to label accurately. RLHF, chain-of-thought reasoning, adversarial testing for generative AI - work that requires someone who actually understands what the model is trying to do.
The strategic vision behind this is worth sitting with: Radha is arguing, and has been arguing for years, that the quality of AI training data is not a commodity problem that gets solved by scale alone. It is a judgment problem that gets solved by expertise. Most of the AI industry is still catching up to this framing. She has been building the infrastructure for it for over a decade.
iMerit operates across healthcare imaging, autonomous vehicles, robotics, natural language processing, geospatial data, and financial services. Their Ango Hub annotation platform - named Best Data Annotation Platform at the Economic Times Making AI Awards 2025 - has become a significant product in its own right. The company has raised $36.3M in total funding, with the most recent round closing in November 2021.
Radha's public profile has grown in proportion to AI's cultural footprint. She writes for TechCrunch, appears at Indiaspora summits and NASSCOM Foundation conferences, gives interviews on responsible AI, and has accepted the kind of awards that get given to people who were right about something important before everyone else caught on. Economic Times named her Impactful CEO in 2025. The Indiaspora Global AI Summit awarded her the She Shapes AI recognition in 2026. Global Thinkers Forum recognized her for excellence in youth development.
What does not come through in award citations but does come through in the company she built: 80% of iMerit's workforce comes from underserved communities. This is not a corporate responsibility footnote. It is the business model. Radha built a company where the social mission and the commercial mission are the same sentence.
