Data will decide winners in generative AI era: Clearwater

With generative AI changing the way enterprises are run, companies should focus on data, which will be a key differentiator, said Sandeep Sahai, chief executive of Clearwater Analytics, a NYSE-listed investment accounting solution provider.

Sahai said people are not thinking enough about where the data is and getting access to the data, as the focus is more on the large language models (LLM). He explained that having the right data to train matters, since businesses require a high degree of accuracy, and makes sure it doesn't hallucinate.

"I shouldn't say this, but most LLMs are somewhat similar in their capability. There are differences and all are quite powerful. Then the only question is, what are you training on? And if you can get the data right in a spot which is readable, guess what? Then you win. That is what we are focused on," Sahai said.


Clearwater Analytics, which manages $9-10 trillion worth of assets and counts companies such as JP Morgan among its clients, has developed its own proprietary model that has been trained on client data in the last two and a half years.

The company recently acquired Enfusion and Beacon, which are in investment management and risk management solutions, respectively, which provide Clearwater Analytics with access to data.

"(With the acquisition) We now have the data about trading, accounting, and risk all at one spot. When you think about Google, all the searches everybody does is in one spot. (Similarly), Whoever has the most data wins, those with access to only public data lose eventually," he said.

Apart from data, one of the biggest challenges are rapidly changing business models. Sahai likened this to the taxi industry and Uber, where instead of making the taxi industry efficient, it had given way to Uber.

He said that with AI, the idea is not to make taxi industry more efficient. "The challenge is really in companies being thoughtful about completely changing. If people think of it like an incremental technology, they will be in trouble," he said.

While Sahai argued that it is naive to declare software is dead, he agreed that generative AI will significantly alter the current state of the software companies. "It is going to make them transform, become more efficient and provide value. (But) the software companies that do not innovate will have a hard time," he said.