Skip to main content
Ekyam’s Batch Integration for Robust Scheduled Operations Business Case: Strategic Data Processing
  • Syncing with Legacy Systems: Several ERPs are designed to generate reports or export data on a scheduled basis (nightly reports). With batch integration, it becomes easy to consume these outputs, thereby allowing modern systems to integrate with and leverage data from these less agile platforms.
  • Bulk Product Catalog Uploads: A batch upload of adding a large number of new products or making extensive updates to an existing product catalogue is more efficient than real-time updates.   This allows for validation and processing of the entire dataset before it is made live. 
Contrast with Real-Time: The Problem of Data Latency The fundamental difference between batch and real-time integration lies in data latency.
  • Real-time integration has near-zero latency, as data is delivered when it is created or updated to give instant inventory updates, prevent overselling and fulfill orders.
  • Batch integration, inherits data latency as data is collected over a period, processed, and then delivered.
Technical Methods: Scheduled Data Movement For implementing batch integration, it involves scheduled data transfers using established protocols: → Scheduled FTP/SFTP File Transfers: Being a common method for batch processing, you need to Source systems export data into files (e.g., CSV, XML, JSON) and place them on an FTP (File Transfer Protocol) or SFTP (Secure File Transfer Protocol) server. The consuming system connects to this server, downloads the files, processes the data and imports it. SFTP is a more preferred method due to its encryption. → Scheduled Database Queries: Batch integration involves scheduled jobs that execute SQL queries to extract large datasets from a source database. This extracted data is then loaded into a destination database, or an application reads it directly from the query results for processing. This method is often used for data warehousing or ETL (Extract, Transform, Load) processes.