The KAI-GPT large language model (LLM) is designed to address the banking industry’s unique needs for accuracy, transparency, trustworthiness and customization
Kasisto, the company behind the digital experience platform for the financial services industry KAI, announced the launch of the banking-specific large language model (LLM) KAI-GPT.
Kasisto’s first generative AI application to be powered by KAI-GPT LLm is the KAI Answers chatbot. It will respond to the queries of bank employees dealing with customer service via a contextual, human-like conversational experience.
The tool is designed to help bankers locate, interpret and understand information from various sources, such as corporate policies, regulatory filings, procedures, web content, and complex financial products.
This bank-specific LLM was trained on a massive amount of financial data and can be used to build customized generative AI applications for banks and financial institutions. Business clients who decide to utilize the service will have “full visibility and auditability of the training methodology and content used to build the LLM,” according to the company’s press release.
Currently, Australian Westpac is reportedly in the process of implementing KAI. The bank has partnered with Kasisto since 2020, exploring how AI may help bring next-generation banking to its customers and employees.
Today, David Walker, CTO of Westpac Group, is “excited to unlock the potential that KAI Answers will bring.” In his opinion, being a banking industry-specific LLM, Kai-GPT is more accurate, safe and intuitive than multi-purpose generative AI applications. At the same time, it’s delivering ChatGPT-like conversational experience, providing human-like, financially literate interactions at speed and scale.
Zor Gorelov, CEO of Kasisto, claims that Kai-GPT technology “is built with the highest standards required by the banking industry.” That presupposes “deep consideration for compliance, accuracy, safety and ethical responsibility,” which is a key challenge for financial institutions in dealing with generative AI tools, that became the most rapidly adopted technology in history.