Vancouver - Recently, Harex InfoTech and its research team, led by Professor Kyoung Jun Lee of Kyung Hee University, won the Innovative Applications of Artificial Intelligence(IAAI) Award at the Association for Advancement of AI (AAAI-24), creating a buzz in the field.
Harex InfoTech's General Commerce Intelligence (GCI) engine was recognized during AAAI-24, which took place over eight days from February 20-27 at the Vancouver Convention Center. The GCI engine stands out as a differentiated technology that delivers AI services while protecting consumer and merchant data. Among the 14 winning projects from around the world, Harex InfoTech's commercial GCI engine was recognized as one of the superior engines that delivers measurable benefits to real-world users of AI application systems.
AAAI is the world’s most prestigious organization in the field of artificial intelligence. Its annual conference brings together leading researchers from around the world to present and discuss the latest research and applications. During the conference, more than 6,000 attendees discussed trends and the future of AI. The event featured keynote addresses by global authorities in AI, including Prof. Yann LeCun of New York University and Meta AI, Francesca Rossi (current president of AAAI) of IBM Research, USA, Prof. Milind Tambe of Harvard University and Google Research, etc.
Prof. Yann LeCun talked about objective-driven AI, which uses predictive world models and self-supervised learning to create machines that can learn, reason, and plan. Prof. Milind Tambe spoke about AI for social impact, which uses machine learning and optimization to address challenges in public health and conservation. Both speakers showed the potential and impact of AI for humanity, as well as the open problems and challenges.
Prof. Kyoung Jun Lee who leads the User Centric AI Research Institute at Harex InfoTech, presented the winning paper titled “General Commerce Intelligence: Glocally Federated NLP-Based Engine for Privacy-Preserving and Sustainable Personalized Services of Multi-Merchants.” This paper garnered significant attention as the first presentation during the conference.
Kyung Yang Park, CEO of Harex InfoTech, stated, “The GCI engine is an AI system that predicts what products or services consumers will buy next. It utilizes cutting-edge AI technology and transformer models.” This cutting-edge engine focuses on analyzing receipt data. However, gathering this data poses challenges, as consumers are hesitant to share their purchase details, and sellers risk revealing sensitive sales information to competitors."
AI Insights Without Data Sharing
Harex InfoTech's UB platform addresses this issue through federated learning technologies. Federated learning, pioneered by Google and Microsoft in 2017, is a new approach to AI learning. With federated learning, individuals, small businesses, and SMEs can develop AI without sharing their own data, and share and federate AI with others to create more powerful AI. Through the GCI engine, small businesses and SMEs can access intelligent services for various marketing and manufacturing needs, including product recommendations, targeted marketing, product bundling, new product ideation, trend analysis, and forecasting.
According to BBQ, the GCI engine is a breakthrough innovation that combines federated learning and natural language processing. It delivers personalized and privacy-preserving AI services to multi-merchant environments. By analyzing receipt data, it generates insights and recommendations for consumers and merchants, while maintaining robust data security.
Existing big tech platforms have made their own social contributions, but they continue to serve as black holes that absorb data from small business owners and SMEs. This may initially benefit small business owners and SMEs, but the problem arises when only big tech platforms gain weight, while small business owners and SMEs lose power over time. As a result, small business owners and SMEs become subservient to big tech platforms, widening the digital gap between these platforms and small businesses.
Consider a scenario where a consumer places an order through a food delivery platform like Uber Eats. Unfortunately, the small business owner lacks visibility into which consumers are purchasing their food. Consequently, small businesses miss out on valuable customer data and lose touch points with their clientele. Meanwhile, large tech platforms have detailed insights into who orders what food, enabling them to optimize food production directly. Ultimately, this discrepancy leaves small business owners relegated to mere mechanical food production
The same scenario applies to e-commerce platforms such as Coupang. In the era of the 3rd Industrial Revolution, this digital platform inadvertently widens the gap. The more consumers, small business owners, and SMEs participate, the greater the divide between big tech platforms and other small businesses and individuals.
To address this issue, UB Platform's GCI engine leverages federated learning technology and combines it with the transformer technology that underpins GPT4 and GPT3.5. This innovative AI technology protects the data of small business owners and SMBs while delivering high-quality AI services. This technology is expected to become a powerful tool in protecting the rights of the digitally disadvantaged through the democratization of AI, and in realizing an AI agent economy where consumers and sellers can transact directly.
UB Platform's GCI engine is used by several franchise companies
The GCI engine is currently being applied to several franchise companies, including BBQ, a big global company, and Ulsan Pedal, a local delivery platform in Ulsan. Through these apps, consumers can receive customized recommendations for new products as well as product bundling based on product type, sales volume, and consumer taste.
Harex InfoTech is also developing an interactive AI ordering system in collaboration with BBQ and other franchises. This system combines generative AI and conversational AI technology to process orders for all goods and services through AI-driven customer conversations.
Prof. Lee, who received the award for UB Platform's GCI Engine, is the fourth recipient of the Innovative Applications of Artificial Intelligence Award in his career. The first award was given to Daewoo Shipbuilding & Marine Engineering in 1995 for developing a specialized scheduling system using AI. The second award was in 1997 for the development of an AI-based process planning system for Hyundai E&C. In 2020, he received the same award for developing an AI system for detecting defective products for Frontec, an SME. The fourth award for developing AI platforms for SMEs, small business owners, and small and medium-sized franchisees is a significant achievement. In 1995 and 1997, the company won awards for developing artificial intelligence systems for large corporations. However, in 2020 and 2024, it was recognized for developing AI platforms for manufacturing and distribution for SMEs.
AI Agent Economic Model Realizes Digital Bill of Rights
Currently, the Korean government has declared the Digital Bill of Rights to protect the rights of consumers and small business owners in the era of digital and artificial intelligence. The goal is to bridge the digital divide and the AI gap with big tech platforms. However, this declaration can only be effectively implemented through a functional AI sharing platform.
CEO Park emphasized that the GCI Engine technology can significantly contribute to the realization of this digital bill of rights. The UB platform built on GCI Engine enables everyone, including small businesses and SMEs, to become AI service providers. By sharing and collaborating on AI with other operators, these individuals can successfully run their businesses without losing customers and data to big tech platforms.
