Kelly Switches to In-House SaaS-Based Delivery System… Achieves Zero Failures
Key point
Kelly completed a full switch from its SaaS-based delivery system to an in-house developed system, achieving zero failures during the transition.
Details
Kelly's Delivery Product Development Team built a new delivery system to address the customization limitations and performance constraints of the existing SaaS-based TMS (Transportation Management System), completing the transition on September 18, 2023. This project focused on enhancing the stability of Saetbyeol Delivery (Dawn Delivery) and improving the efficiency of large-scale data retrieval and monitoring.
Architecture and Transition Strategy
The new system operated concurrently with the existing system, adopting an approach that received the same Kafka topics and filtered them independently. This was a strategy designed to minimize the scope of changes and enable independent response during rollback without assistance from other departments. Additionally, the team utilized Feature Toggle to control whether the new system was applied by region, expanding it in stages.
Phased Transition and Results
The transition process proceeded through 5 stages, from small-scale regional testing to 100% application across all centers, managing real-time issues through an on-site response system. As a result, the transition period achieved zero undeliverable cases and zero failures, a record considered unprecedented in Kelly's project history.
Organizational Culture and Lessons Learned
The key factors behind the project's success were the shift from a functional organization to a strengths-based collaborative structure based on Squads, and the incorporation of insights from on-site experts such as Kelly Next Mile into the design. Going forward, the team plans to explore ways to balance early MVP-level releases with operational burden in order to reduce uncertainty.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.