KakaoBank Hosts KodeRunner 2026, Presenting Agentic AI Vision Centered on 'Reliability'

KakaoBank held its internal technical conference KodeRunner 2026 at the Grand Ballroom of COEX in Seoul. The core message of this event was Reliability, focusing on the safety nets and fundamental trustworthiness that financial banks must establish during the transition to Agentic AI.

Agentic AI Transition and Reliability Assurance Strategy

In the keynote, CEO Daniel emphasized that banks must evolve from being 'deposit and withdrawal-centric' to becoming the 'No. 1 Gateway', an essential 'financial tool' in customers' lives. He presented a vision of 'Agentic AI-based business innovation' where AI Agents automate the entire process from planning to deployment, with AI serving as a tool to assist in risk management and core decision-making. CTO Conrad proposed 'Agent Mandate', where cross-team collaboration is executed by AI Agents to resolve productivity bottlenecks. As core prerequisites for this, he presented the following three strategies for ensuring reliability:

• Data Reliability: Providing verified, up-to-date data (AI-Ready Data)
• Procedural Reliability: Control within regulations and governance, with human intervention in exceptions (AI-Ready Workflow)
• System Reliability: Secure Runtime, Agentic API, Observability (AI-Ready Foundation)

Global Partner Collaborations and Session Structure

Four global Big Tech companies—AWS, Microsoft, Google, and OpenAI—led hands-on labs, while a total of eight partners, including JetBrains, Snowflake, Datadog, and Dell, participated in booth operations. One-on-one hands-on guidance was provided by current engineers from Microsoft (GitHub Copilot App), Google (ADK 2.0), AWS (Bedrock AgentCore), and OpenAI (Codex). Additionally, technical sessions sharing failure and troubleshooting cases in production environments were held, and a new Soft Skill Track covering mental management and neuroscience was introduced.

Event Outcomes

The event attracted over 900 attendees, the largest turnout to date, with an overall satisfaction score of approximately 4 out of 5. 92.42% of participants attended technical or hands-on sessions, and the session on 'Real Production Failures and Overcoming Cases' was the most popular, achieving a satisfaction score of 4.22. The grand prize of the internal AI innovation competition 'The Challenge' was awarded to the 'BankPlay' team for their emulator-based app testing integration platform.