A global brand in high-end electronics had successfully implemented a B2C and B2B digital commerce presence, based on SAP hybris. However, consumers and dealers reported issues and demanded improvements.
The client had a central digital-commerce team responsible for managing requests from their various countries and channelling development work to a near-shore development team in Eastern Europe. Although experienced in development, the team lacked expertise in other areas, such as business analysis, testing, infrastructure, technical operations, DevOps, and automation.
We were asked to implement a framework for operating the platform. We also were to include a next-generation Agile process with collaboration between global teams, a professional in-sprint testing process, test automation, continuous integration, deployment to five test environments and production, and an application-maintenance team.
This project’s specific organisational and technical challenges included:
- Critical test cases were undefined
- Weak collaboration between the Eastern European and Asian teams
- A high number of production hotfixes, which the development team considered to be the client’s problem
- The hosting party wasn’t collaborative
We completely optimised the development and delivery pipeline. The new hosting platform offers better response times and higher service levels.
First, we on-boarded a support team. An experienced test manager from our India office spent time with the client, documenting critical flows and test cases so their team could perform regression testing. We started automating tests, adding more coverage with every sprint. Due to the budget, capacity was limited. After 18 months, we had automated around 85% of the critical test cases.
We helped our client migrate to a different hosting platform. We designed the target architecture, and managed the whole project and the transition.
Initially, only minor bug fixes were required, but the scope and complexity of the work increased. The development team in Eastern Europe used a Scrum approach with 2-week sprints. The application-maintenance team applied Kanban for extra flexibility.
Two different yet complementary development teams meant greater flexibility. As a result, the development and delivery pipeline was completely optimised, with standards and guidelines for defining work packages, based on high fidelity tooling.
As part of the Continuous Integration (CI) pipeline, we checked many aspects of code quality and automated deployments. Plus, with regression testing largely automated, a full regression test now can be done within a day (including reporting). Any release delivered at the end of a sprint can go into production within a few days.
The new hosting platform is more agile and responsive. This means that we produced better response times and higher service levels. If there are any production issues, the operations support team can start root-cause analysis immediately, with client updates on the chat channel.
When there’s an urgent development task, the product owner can instead give it to the application-maintenance team to execute in Kanban mode. There’s also flexibility to add tasks when needed.