AI & Enterprise
Finance AI bottleneck is data, not model performance, solved with ontology
Han Tae-kyung (한태경), head of the AI stock market data vendor business at Weaker Korea, said the bottleneck in finance AI lies in data design, not model performance. He said finance AI is limited without an ontology that links on-the-ground expertise, proprietary data and analysis tools to large language models. Han said generative AI has boosted productivity in tasks such as earnings analysis and topic extraction, but up-to-date data connections matter most in practice.
