•1 min read•from InfoQ
DoorDash Uses Multi Agent LLMs to Clean up 60,000 Feature Flags
Our take
DoorDash has demonstrably streamlined its engineering workflows by deploying a multi-agent LLM system to manage its extensive feature flag landscape—over 60,000 flags across 623 repositories. This innovative approach, integrating live experimentation data and automated validation, achieved a remarkable success rate: 45 out of 50 flags yielded usable pull requests, averaging just 13.8 minutes and $4.79 per cleanup. This highlights a powerful shift towards automated engineering, echoing themes explored in "The Death of the Button," which examines the evolution of user interfaces.
DoorDash built a multi-agent LLM system to automate stale feature flag cleanup across more than 60,000 flags and 623 repositories. The workflow combines live experimentation data through MCP, engineer approval, isolated Git worktrees, parallel agents, and automated validation. In an evaluation of 50 flags, 45 produced usable pull requests at an average of 13.8 minutes and $4.79 per cleanup.
By Leela KumiliRead on the original site
Open the publisher's page for the full experience