The company is introducing its VUSSENS platform to the North American market, leveraging a specialized LLM architecture for cross-channel trend analysis.
NEWEN AI will demonstrate its VUSSENS platform at Cosmoprof North America Las Vegas 2026, scheduled for July 13 to 15, marking a strategic expansion into the North American beauty sector. According to a company press release, the platform utilizes a proprietary Beauty AI Ontology to map complex relationships between consumer sentiment, ingredient efficacy, and market performance metrics within a highly competitive regulatory environment.
At the architectural core lies the QuettaLLMs-27B-Koreasoner-V3, a large language model that recently secured the top position on South Korea’s K-AI Leaderboard. This model processes over 800 billion tokens of domain-specific training data to facilitate high-granularity persona analysis across diverse demographic and physiological variables, ensuring alignment with regional market standards.
The system integrates a Multimodal AI Engine designed to interpret unstructured inputs, including video, audio, and text, to capture real-time market signals. By correlating short-form video engagement with purchase conversion data, the engine identifies causal relationships within the beauty supply chain, providing actionable insights for local distributors.
The platform architecture comprises four distinct functional modules: Marketing, Trend, Product, and Category. These modules operate on an integrated data pipeline that tracks seeding status, content engagement, and real-time reaction metrics to ingredients and product textures, enabling a comprehensive view of market dynamics.
Technical validation of the underlying framework has been conducted through partnerships with global enterprises including L’Oréal, Amorepacific, and Cosmax. These collaborations have allowed for the refinement of the proprietary Quetta architecture, which currently supports over 550 industrial projects across various sectors.
The company maintains a significant focus on data-driven decision support for ODM companies and retailers. By leveraging the NYU Stern MBA network and the Global AI Frontier Lab, NEWEN AI aims to standardize its analytical approach for the North American market, supported by the 2026 Industrial Voucher Program from the Ministry of Trade, Industry and Energy.
The reliance on a domain-specific ontology rather than general-purpose LLM architectures allows for greater precision in identifying niche market trends. This structured approach to data ingestion mitigates the hallucination risks often associated with broad-spectrum generative models in specialized industrial applications, ensuring data integrity.
The integration of multimodal inputs addresses the limitations of traditional text-based sentiment analysis in the beauty industry. By quantifying visual and behavioral data, the platform provides a high-fidelity predictive capability for identifying emerging consumer preferences before they reach mass-market saturation.
The scalability of the Quetta architecture suggests potential for future expansion into other sectors beyond beauty, such as fashion or retail. The current focus remains on establishing a Global Beauty Data Hub to facilitate cross-regional trend analysis and benchmarking, supported by the K-AI Leaderboard performance metrics.
The underlying QuettaLLMs-27B-Koreasoner-V3 model utilizes a specialized training methodology that prioritizes domain-specific tokenization. This ensures that the model maintains high performance when processing technical ingredient data and complex efficacy claims that often confound standard language models, providing a distinct advantage in R&D.
Data ingestion pipelines within the VUSSENS platform are designed to handle high-velocity streams from social media and e-commerce platforms. By normalizing this data into the Beauty AI Ontology, the system ensures that disparate signals are comparable across different geographic regions and consumer segments for better decision-making.
The platform’s ability to perform SWOT analytics on a per-ingredient basis represents a significant advancement in automated market research. By automating the extraction of efficacy signals from unstructured reviews, the system reduces the manual labor required for product development cycles and enhances strategic planning.
Industry observers should monitor the performance of the VUSSENS platform as it undergoes real-world testing in the North American market. The upcoming deployment will serve as a critical benchmark for the efficacy of the company’s multimodal approach in high-velocity consumer goods environments, potentially setting a new standard for AI-driven market intelligence.
