"The transition to artificial intelligence (AI) in businesses is not merely about adopting AI tools; it requires a comprehensive redesign of work processes, data management, and organizational operations," said Shin Gye-young, Vice President and head of the AI Business Team at Samsung SDS.
Shin made these remarks during a presentation on 'Transitioning to AI Native Enterprises' at the Samsung SDS AX Summit held on May 29 at the company's Jamsil campus in Songpa, Seoul. The summit focused on sharing strategies and real-world applications of AI agent-based AX (AI Transformation).
He noted, "There is a saying that AI does not take jobs away, but rather those who effectively utilize AI may take jobs from others. Similarly, if companies do not properly implement AX, they risk falling behind in competition."
Shin defined AI native transition as the innovative redesign of overall business and IT processes based on AI. He emphasized the need for a hybrid strategy that combines both top-down and bottom-up approaches for implementing AX.
The bottom-up approach involves enhancing productivity and AI literacy through the use of AI at the employee and departmental levels, although it can be challenging to measure the overall impact of investments. In contrast, the top-down approach focuses on redesigning entire workflows, allowing for clearer assessment of key performance indicators (KPIs) and return on investment (ROI).
To measure the success of AX, Shin proposed three key indicators: AI performance KPIs, workflow KPIs, and business KPIs. He explained that it is essential to evaluate how much AI has improved accuracy and coverage compared to existing tasks, reduced work hours and personnel input, and ultimately led to improvements in quality and productivity.
Shin identified six core pillars for AI native transition: process redesign, securing AI-ready data, developing and operating agents, governance, talent development, and fostering an organizational culture. He particularly highlighted the importance of data cleansing, metadata management, and access control.
The operational framework for agents was also noted as a key challenge. Shin stressed the need for environments where general employees can create agents using natural language, as well as low-code development environments that operate according to workflows, alongside traditional coding environments for developers. He explained that created agents should be registered in an internal marketplace for reuse, with integrated management of permissions, usage, token costs, and risks associated with external tool usage.
Samsung SDS shared its own organizational practices, stating that it established a structure to lead the company's AX transition through a reorganization last year, encouraging the use of AI tools across various departments. Currently, about 107 AI professionals are actively engaged, and the company has implemented systems for identifying best practices and providing training and certification.
Customer examples included Woori Bank and Samsung Electronics. Woori Bank announced its transition to AX, designing over 175 agents across five major functions and 27 core tasks. Samsung SDS has built more than 300 agents based on its 'Fabrics' platform, supporting the redesign of financial operations such as customer management, credit, and asset management.
To support AX, Samsung SDS is integrating its data analytics platform 'Brightics AI,' agent platform 'Fabrics,' and automation platform 'Brite Automation.' Additionally, the company is facilitating the adoption of OpenAI Enterprise to assist domestic companies in expanding their use of AI.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.