The customer, a leading fintech company that provides payment services, has been committed to enable public sector organizations and corporations to realize operational excellence. The customer’s legacy data platform was on Oracle and faced several performance issues, which in turn impacted the SLAs. The customer partnered with Blazeclan for modernizing its data ecosystem to bring scalability and performance improvements.

The Need for Modernizing the Data Ecosystem

The customer’s existing data platform, including the code and storage, was based on Oracle. This resulted in manual maintenance of code and infrastructure and being a proprietary tool, it added to cost. The on-premise data ecosystem faced performance issues and scalability vis-a-vis growing demands was challenging.

Key requirements of the customer:

  • Modernizing the data platform and resolving the issue of technical debt of old legacy code.
  • Preventing SLA-related issues resulting from performance bottlenecks.
  • Automating the platform maintenance and enhancing scalability.

The Solution

Blazeclan proposed to the customer a solution that would provide them with a completely re-architected platform on AWS cloud. The solution enabled the customer to achieve multi-tenancy in their architecture and facilitated application maintenance.

The Approach

  • Converted the Oracle stored procedures to Apache Spark-based code, in turn achieving faster-distributed processing.
  • Implemented AWS EMR for Apache Hadoop, which helped the customer achieve an elastic and scalable infrastructure.
  • Used transient clusters for storage with Amazon S3 to enable cost optimization.
  • Re-architected the entire data platform on the AWS cloud to resolve all scalability and performance issues.

Architecture Diagram

Benefits Achieved by the Customer

  • The AWS cloud-based solution enabled the customer to achieve easier and quick scalability based on workloads..
  • The new cloud-based data platform resulted in significant cost optimization for the customer, with effective use of transient clusters.
  • The solution was implemented on a reusable framework. Also, the solution resulted in faster and easier research & development activities. 
  • The data platform was free of proprietary databases, which enabled cost reduction.