Isotopes emerged from stealth Thursday with a $20 million seed round, positioning itself to solve big data's most persistent problem: bridging the gap between those who manage data infrastructure and those who actually need to use it. Led by Hadoop co-creator and former Scale AI CTO Arun Murthy, the startup promises its AI agent can finally democratize enterprise data access through natural language queries.
Isotopes just threw down the gauntlet in enterprise AI with a $20 million declaration that it can solve big data's oldest problem. The startup emerged from stealth Thursday with an AI agent designed to bridge the decades-old chasm between data engineers who build systems and business users who need insights.
The pedigree behind this ambitious claim runs deep. Co-founder and CEO Arun Murthy helped create Hadoop over two decades ago at Yahoo, sparking the initial Big Data revolution of the 2010s. After co-founding Hortonworks in 2011 and taking it public four years later, Murthy witnessed firsthand how even data companies struggled with data access. "It was embarrassing," he told TechCrunch. "We were a big data company selling this" yet couldn't answer Wall Street analysts' questions during earnings calls.
That frustration led Murthy to Scale AI in 2021, where he served as CTO under founder Alexandr Wang. The experience was "like getting a PhD at Scale," Murthy explains, giving him deep insights into what drives AI models and how to improve them. But when former Hortonworks colleague Prasanth Jayachandra reached out about starting their own venture, the opportunity proved irresistible.
Isotopes launched in late 2024 with a third co-founder, Gopal Vijayaraghavan, also from the Hortonworks era. Their seed round came from NTTVC's Vab Goel, formerly of NorWest Ventures, validating the founders' vision for reimagining enterprise analytics.
The company's AI agent, called Aidnn, goes far beyond simple chatbots that have flooded the market. When business managers ask questions in natural language, the agent doesn't just search existing data—it orchestrates complex multi-step processes to create the insights they need. "The data that you want to chat with actually doesn't exist, at least the form that you need to chat with," Murthy explained to TechCrunch. The agent executes sophisticated workflows: extracting metadata, cleaning and normalizing information, joining datasets, prorating revenue, and aggregating results.
