Cloud-based supply chain traceability | Legacy system modernization

Challenge

Mainframe legacy systems were not scalable to the rising volume of part sales and repairs.
Attempt to translate and rewrite Cobol to Dotnet was unsuccessful due to spaghetti code in legacy.
Converting data to API using IBM DataPower for front end to consume resulted in drastic performance issue at tables.
CDC products or oracle golden gate solutions were adding huge license cost on top of existing spend.

Solution

Build a custom REST API solution on top of existing mainframe database using Apache Camel, Spring, Jetty.
Apache Camel applied transformations required for API.
Cached datasets and used ETags to optimize calls to database saving performance.
Exposed REST API over IBM data power to external systems to consume.
Microsoft Azure based cloud native apps synced data and rendered over PowerBI for analytics and predictions.

Benefits

Solution didn’t require any new software licenses.
Mainframe database capacity was preserved well for other legacy processes.
No legacy system update was required with growing sales, suppliers, plants and dealerships.
The cloud native solution was able to perform recall predictions much earlier.
End to end traceability of parts.