The era of extensive artificial intelligence (AI) experiments is now shifting to a new phase. In a m..

Source Domain: www.mk.co.kr

사진 확대 Lynsey Patel, Executive Vice President, Snowflake
The era of extensive artificial intelligence (AI) experiments is now shifting to a new phase.

In a market with a solid digital infrastructure and a clear regulatory system like Korea, discussions are unfolding beyond the simple introduction of AI. Commercial responsibility is required at the next stage of AI maturity.

Financial institutions have reached a stage where they have to actually bring out improvement in customer retention, profit generation, and operational efficiency through AI. AI initiatives are now being evaluated based on whether they can produce results that can be measured on a large scale.

The financial services industry operates in a highly regulated and data-intensive environment. In order to achieve meaningful return on investment (ROI) in such an environment, general-purpose AI solutions alone are not enough. The most effective system is one that has accumulated institutional-specific knowledge over decades, including credit risk cycles, regulatory requirements, long-term customer behavior, and operational history. At this point, many early AI initiatives reveal their limitations. Without access to complete and contextualized data, no matter how advanced the model is, it gives only fragmentary insights into risks, opportunities, and customer behavior.

Successful AI strategies are essentially data strategies. Many financial institutions are experiencing problems in which customer information, transaction records, and risk models are distributed in different systems. When data is individualized, AI only provides a partial view of the business and misses important insights because it cannot see the entire picture.

Liquidity risk management is a case in point. Meeting regulatory requirements requires the integration of data from multiple sources under strict time constraints.

Integration of risk, financial, and market data…

Source