Leading the implementation of changes in the existing Amazon e-commerce platform to enable the launch of the Amazon Prime grocery business. Amazon launched an ambitious new business delivering groceries to households within 30 minutes of order placement. Multiple local nodes served customers within a 10-kilometer radius to meet business KPIs. These nodes were called fulfillment centers (FCs). Based on weekly order predictions, the inventory requirements for various SKUs were estimated, and these products were transshipped from the distribution center (DC).

The existing Amazon app, catalog system, order processing system, delivery system, transshipment system, payment system, and return and disposal system were modified to suit grocery use cases. Nodes were launched in a phased manner, data was collected, and customer behavior and order placement patterns were analyzed. Based on these insights, the systems were further modified. This iterative mode of development helped us understand customer acquisition, order conversion, and retention.

The biggest challenge in the program was making changes without impacting the main Amazon.com platform. To overcome the possible risk of introducing defects or issues in the Amazon platform system, the following processes were followed

A comprehensive list of all P0, P1, and P3 use cases was documented, detailing the required changes for the main app and website. The use cases were reviewed and approved by business and operations owners from both business lines.

System Documentation: Detailed component-level information, impact analysis, and required changes were documented.
Traceability: Use cases were mapped to system-level change documents to ensure complete traceability and coverage.
Code Changes: Low-level code modifications were documented, along with the relevant code branch SHA.
Risk Management: Risks and assumptions were identified, with corresponding mitigation and contingency plans.
Collaboration: All documents were shared and reviewed with team owners across the US, Europe, Japan, and India.
Resource Allocation: Teams identified and assigned resources needed for the changes.
Execution Planning: Detailed dependencies, an execution plan, and completion timelines were established for each system.
Progress Tracking: Weekly progress tracking assessed risks of delays and business impacts, prompting corrective actions when necessary.
Launch Preparation: Node launch steps were documented, and trial runs were executed on sample nodes.
Go-Live Monitoring: A launch team, including POCs from each team, was assigned to oversee the go-live process, post-launch KPIs, and corrective actions.
Future Development: Comprehensive steps, processes, guidelines, team POCs, and timeline estimations were documented to guide future platform development