Learn about Oracle Machine Learning for Oracle Database and Big Data, on-premises and Oracle Cloud

Machine learning helps financial company grow business

Case Study: Forth Smart Corporation uses Oracle Database and built-in Oracle Machine Learning to explore new opportunities.


About the company

Forth Corporation is based in Bangkok, Thailand, and has its subsidiary, Forth Smart Corporation, operating more than 120,000 kiosks nationwide for providing financial services like recharging prepaid mobile phones, bank transfers, utility bill payments, and paying for mobile games. Recently they have also expanded into offering vending machines for snacks, drinks, and coffee. Thailand is primarily a cash-based economy and the majority of the population does not have credit or debit cards.

Forth Smart Corporation works with 100 service agent partners who collect cash from the 120,000 kiosks nationwide, deposit in the bank, and transfer to the corporate account. They are also responsible for servicing the kiosk machines during power outages or any damage. There is a support team of more than 40 people to help these 100 agents.

How Oracle supports Forth Smart Corporation

Kiosks are connected to Oracle Database, and Oracle Autonomous Database via Oracle GoldenGate Cloud Service to sync data in real-time across the databases, allowing agent service dashboards to run in real-time. The solution also includes Oracle Data Miner, a GUI-based drag and drop interface to very quickly build and deploy ML models in the database, and Oracle Analytics Cloud to build dashboards for kiosks observability and problem detection in real-time.

machine learning architecture for financial services

Forth Smart Corporation uses in-database Oracle Machine Learning to increase its scope and explore new opportunities, using existing Oracle investments to build, evaluate and deploy ML models. With a growing business, Forth Smart Corporation needed to scale its services and quickly provide solutions to its customers who are retailers providing services to end consumers.

Line is a consumer app that sells emoticons at $1 or $3, with a large portion of its customers who are cash-based. Forth Smart Corporation was able to query its 15 million customer-base actively using its Oracle Cloud database. Using learning models on this dataset, they were able to query a target customer base for Line that led to a 3X conversion rate and an even greater 6X conversion rate by targeting these groups on a specific social media platform instead of SMS messaging.