go to news

detail

* It has been translated using AI
Minjoo Kim
Input : 
2026-05-18 15:33:40
BC Card was invited as a presenter for the first time among domestic financial companies at the "Red Hat Summit 2026" held in Atlanta, USA from the 11th to the 14th, and introduced an example of the "Agentic AI Operation Platform" built with Red Hat. The photo shows Lee Tae-young, head of BC Card AX team, presenting actual service application cases at a local event. [BC Card]
BC Card was invited as a presenter for the first time among domestic financial companies at the "Red Hat Summit 2026" held in Atlanta, USA from the 11th to the 14th, and introduced an example of the "Agentic AI Operation Platform" built with Red Hat. The photo shows Lee Tae-young, head of BC Card AX team, presenting actual service application cases at a local event. [BC Card]

BC Card announced on the 18th that it announced an example of an 'Agentic Artificial Intelligence (AI) operating platform' at the 'Red Hat Summit 2026, an open-source technology conference.

The Red Hat Summit is the world's largest open-source, cloud, and AI technology conference held annually by global open-source company Red Hat in New Orleans in 2005, and is a global event attended by tens of thousands of people including IT leaders, developers and engineers from around the world every year.

BC Card unveiled for the first time on the global stage on the 14th (local time) an example of the 'Agentic AI operation platform' jointly built with Red Hat over the past year.

It is the first time that a domestic financial company has been invited to the event as a presenter.

Agentic AI refers to AI that judges and acts on its own beyond simple answers. Using this, the 'Agentic AI Operation Platform' allows a number of small language models (SLMs) specialized for each area to be swarmed and processed simultaneously in the existing method that relied on all tasks for one large language model (LLM).

In particular, for LLM, there were limitations of high operating cost and slow processing speed, but SLM was optimized for certain tasks, which allowed for stable results as well as improved inference speed by up to 3 times, while also reducing GPU resources by up to 70%.

At the event, BC Card announced an example of efficiently distributing expensive GPU resources by integrating its accumulated know-how by processing more than 10 billion payments data annually with AI technology. In the process, operating costs were drastically lowered and service speeds were more than tripled.

Beyond simply introducing AI services based on these technological achievements, ▲Eat.It also introduced services such as pl (recommending restaurants after analyzing real-time consumption patterns) ▲ BCGPT (support for executives and employees' work through financial-specific Generative AI platforms) ▲ MOAI (implementation of automation of work by connecting multiple AIs) and attracted interest from global companies.

In addition to developing R&D technology, various activities are also being conducted to contribute to the development of global AI infrastructure.

BC Card has released about 38 LLMs as open sources on 'Hugging Face', a global LLM open-source hub, so that developers around the world can use their own AI models and Korean financial-specific datasets.

Through this, it is also the only global LLM country viewer in the domestic financial sector and has achieved tangible results with an average of 100,000 downloads per month. This is evaluated as a representative example of taking off the image of the closed financial sector and showing its appearance as a leader in leading technology standards around the world.

Oh Sung-soo, head of BC Card's AI Data Division, said, "The announcement at the Red Hat Summit was a meaningful place where BC Card's global-level Agentic AI engineering capabilities were officially recognized on the world stage. We will continue to show our identity as a technology entity contributing to the open-source ecosystem beyond just introducing AI on behalf of the Korean financial sector."

#

Most Read News