Ex-Scale AI CTO Launches AI Agent to Solve Big Data’s Biggest Problem
▼ Summary
– Isotopes AI emerged from stealth with a $20 million seed round and offers an AI agent called Aidnn to bridge the gap between data infrastructure experts and business users.
– The agent allows managers to query data in natural language, generate complex planning documents, and pull information from various sources like finance apps, ERP, and CRM systems.
– Co-founder CEO Arun Murthy has a strong background in big data, having co-founded Hortonworks after contributing to Hadoop, which fueled the initial big data boom.
– The agent distinguishes itself by cleaning data, maintaining context memory, and handling multi-step processes like data extraction, normalization, and aggregation for complex tasks.
– Despite its sophistication and patent applications, Isotopes faces competition from incumbents like Salesforce’s Tableau and other startups in the AI agent analytics market.
Isotopes AI has officially launched with a substantial $20 million seed investment, introducing an advanced AI agent designed to bridge the long-standing gap between data infrastructure specialists and the business users who rely on that data. This new platform, named Aidnn, enables managers to interact with their data using natural language queries, pulling information from diverse sources including finance applications, ERP and CRM systems, and cloud storage. By automating data retrieval and even generating detailed planning documents, Isotopes aims to solve one of the most persistent challenges in enterprise analytics.
The company stands out not only for its technology but also for the impressive background of its founders. Arun Murthy, Isotopes’ CEO, brings decades of experience from the heart of the big data revolution. Two decades ago, he contributed to the development of Hadoop at Yahoo, a project that ignited the big data movement. He later co-founded Hortonworks, which went public within just four years. Despite industry shifts that eventually led to Hortonworks merging with Cloudera, Murthy continued to observe a critical problem: even within leading data companies, decision-makers often lacked timely access to the information they needed.
Reflecting on his time at Cloudera, Murthy recalls quarterly calls where executives struggled to answer analysts’ detailed operational questions simply because the data wasn’t accessible. “It was embarrassing,” he admits. “We were a big data company selling this.” After leaving Cloudera in 2021, Murthy joined Scale AI as CTO, where he deepened his understanding of AI model behavior and improvement, an experience he describes as akin to earning a PhD in modern AI systems.
The idea for Isotopes took shape when Murthy reconnected with Prasanth Jayachandra, a former colleague from the Hortonworks era. Together, they brought on Gopal Vijayaraghavan, another veteran from those early big data days, to form the founding team. Late in 2024, they officially established Isotopes, securing seed funding led by Vab Goel at NTTVC.
What sets Isotopes apart in a crowded field of AI analytics tools is the depth of its agent’s capabilities. Aidnn doesn’t just retrieve data, it cleans, normalizes, and contextualizes it, performing multi-step operations that traditional chatbots cannot handle. For instance, when tasked with drafting a report on monthly recurring revenue trends, the agent must extract metadata, clean and join datasets, prorate revenue, and aggregate results, all while documenting its reasoning and flagging anomalies.
Murthy emphasizes that this is “far beyond a simple chatbot.” The platform also prioritizes data privacy, allowing enterprise clients to deploy the agent without sharing sensitive information with external AI model providers. Despite these advanced features, Isotopes enters a competitive landscape. Established players like Salesforce’s Tableau are rolling out their own AI agents, and other well-funded startups such as WisdomAI are also vying for market share. Yet with its unique founding DNA and patented technology, Isotopes is positioning itself as a sophisticated solution for organizations tired of waiting for answers buried in their own data.
(Source: TechCrunch)
