{"slug":"unlocking-the-potential-of-social-banking-to-the-underserved","url":"https://findausecase.com/use-cases/unlocking-the-potential-of-social-banking-to-the-underserved","title":"Unlocking the potential of social banking to the underserved","description":"KASIKORN LINE, operator of Thailand's LINE BK social bank, used the Databricks Data + AI Platform to build a credit scoring model that combines conventional financial data with behavioral data from the LINE messaging app to assess creditworthiness for underbanked and unbanked customers. The platform cut time to insights by up to 60x, scaled data volumes for training ML models 2-5x, and accelerated experimentation for social banking models by up to 2x.","company":"KASIKORN LINE","industry":"Financial Services","country":"Thailand","aiCapabilities":["Predictive Analytics","Machine Learning"],"technology":["Databricks Data + AI Platform","Delta Lake","Databricks SQL","Unity Catalog"],"deployment":"Public Cloud","problemStatement":"KASIKORN LINE encountered challenges managing data silos, diverse sources and real-time credit decision-making while dealing with vast volumes of social and financial data. As a lean team, the massive data volumes made it crucial to have a scalable and integrated platform to unlock actionable insights, serve customers better and drive financial inclusion.","solutionApproach":"KASIKORN LINE adopted the Databricks Data + AI Platform, a lakehouse unifying data with BI analytics, reporting and ML capabilities, to develop a credit scoring model that combines conventional financial data with behavioral information from LINE to evaluate creditworthiness for underbanked and unbanked customers. Delta Lake enhanced data pipeline reliability while reducing latency, Databricks SQL let teams run ad hoc queries with ease and precision, and Unity Catalog provided secured, granular access control for sensitive financial data across teams in different countries.","businessValue":"Complex data analysis that once consumed hours has been reduced to minutes or seconds, slashing time to results by up to 60x. KASIKORN LINE has also scaled the data volumes used to train ML models by 2 to 5x and accelerated model experimentation by up to 2x.","evidence":{"band":"high"},"sourceUrl":"https://www.databricks.com/customers/kasikorn-line","dates":{"publishedAt":"2026-09-22T05:48:07.794Z","publishedAtSource":"pipeline","updatedAt":"2026-09-22T05:48:07.794Z"},"license":"Open for reading and citing with a link to https://findausecase.com/use-cases/unlocking-the-potential-of-social-banking-to-the-underserved. Bulk republication requires permission."}