THK overhauled its core data integration platform and established stable data integration for approximately 1,100 tables from IBM i to Snowflake, reducing synchronization time for large-scale data from several hours to several minutes.
THK migrated its data integration platform to Snowflake and introduced Qlik Replicate and Qlik Enterprise Manager, establishing a new framework for integrating data from its global core systems into Snowflake. Although synchronizing large volumes of data took several hours in the legacy environment, the overhaul has enabled approximately 1,100 tables to be synchronized reliably.
A proof of concept based on the production environment was conducted for approximately 3 months to verify solutions to technical issues specific to IBM i, including character encoding conversion, file locking and timeout settings. The integration method was also optimized with consideration for data freshness, consistency and cost, with update frequency adjusted according to the intended use.
Operational workloads were significantly reduced through centralized management and automated monitoring using QEM. Tasks on 4 Qlik Replicate servers are centrally managed and monitored, and a notification framework for anomalies has been established.
Going forward, THK aims to create an environment in which generative AI can be used to access internal information in natural language, with data on Snowflake as the starting point. The company plans to develop the platform so that the sales, factory and head-office divisions can use the data they need for decision-making.