Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore SK Telecom's journey in developing LLM-based recommendation systems through their comprehensive 'One-Model' project evolution from 2023 to present. Learn how SK Telecom transformed their fragmented, service-specific recommendation systems into a unified enterprise-wide solution that leverages Large Language Models to provide personalized recommendations across telecommunications services including rate plans, T-universe subscriptions, memberships, Tworld AI, A.Dot, and Tdeal. Discover the technical progression from Transformer-based architectures in versions 1 and 2 that integrated multiple domains, to version 3's LLM backbone approach that solved cold-start problems, culminating in version 4's generative LLM recommendation model that goes beyond simple recommendations to provide reasoning, marketing messages, and related product suggestions. Gain insights into the practical challenges of implementing LLM-powered recommendation systems at scale, including domain integration strategies, cold-start problem solutions, and the transition from traditional collaborative filtering to generative AI approaches. Understand the business impact and technical achievements of this project, including recognition at international conferences with SIGIR 2024 Best Paper Honorable Mention for version 2 and CIKM 2023 paper acceptance for version 1. Learn from SK Telecom's AI research engineers Kim Tae-san and Yoon Hyung-jun as they share real-world implementation experiences, technical obstacles overcome, and measurable service improvements achieved through their LLM-based recommendation system deployment across SK Telecom's customer-facing services.
