Testing, evaluation and compliance expected to form 25% of entire AI market by 2030: RagaAI’s founder

By the end of this decade, artificial intelligence (AI) systems testing, evaluation, and compliance will form 25% of the entire AI market, predicts Gaurav Agarwal, the founder and chief executive officer (CEO) of RagaAI.

“Research reports indicate that by 2030, the AI market value is poised to grow to $2 trillion. Our estimate and analysis suggest that 25% or $500 billion of this value will come from testing and evaluation tools and services,” Agarwal told TechCircle. RagaAI, founded in 2022, offers a testing platform and solutions for AI system and tools. In January, the company raised $4.7 million in seed funding in a round led by pi Ventures, with participation from investors such as Anorak Ventures, TenOneTen Ventures, Arka Ventures, Mana Ventures, and Exfinity Venture Partners.

An industry veteran and an ex-Ola Electric and Nvidia employee with specialisation in the autonomous driving space, Agarwal said that while there has been a lot of work happening in AI development, inadequate testing causes issues during deployment and mass scaling up. The idea for RagaAI was conceived after Agarwal suffered a potentially life-threatening ordeal. “Back in the time when I was working with Nvidia in San Fransico, I was test-driving a semi-autonomous car. I was driving on a highway at a high speed. The car was supposed to spot the debris ahead and slow down. It was in a split second, I identified this failure and manually applied brakes. This was a scary incident which got me thinking about how much of such failures can hold back AI’s immense potential,” he said. 

Testing and evaluation of AI has been an ad-hoc, best practice effort; it is not comprehensive and inadequate, claims Agarwal. “In summary, our solutions have helped customers reduce errors by 90% and accelerate time to market by three times for our customers, which is a significant value add.” With the seed fund, RagaAI, which currently operates in India and the US, will expand customer outreach, along with sales and marketing efforts. The company will also be investing in new innovation and expand technology stack. The company, at present, offers 400 out-of-box tests for AI. “We have built a horizontal platform which can support a wide variety of use cases.” Last month, RagaAI open-sourced RagaAI LLM Hub, its enterprise-ready large language models evaluation and guardrails platform. 

Agarwal believes that the demand for platforms such as RagaAI will grow in the future, given the AI compliance and regulatory requirements enforced by governments around the world. To be sure, earlier this month, the United Kingdom and the United States signed a joint agreement on advancing testing and evaluation of AI technologies. Last month, the ministry of electronics and IT (MeitY) under the government of India released an advisory that asked intermediaries to seek government approval before deploying AI models. After much criticism from the community, the government revised the advisory, removing the approval part but asking firms to label under-tested models.