Netflix Reworks Conductor for 420 Million Monthly Workflow Executions and 10X Larger Workflows
Netflix scaled Conductor to 420M monthly workflows. The move: separate metadata from tasks, async evaluation, dynamic workers. 10x larger pipelines now possible.
Why it matters
Netflix's Conductor 4.0 demonstrates production-scale workflow orchestration at massive volume — relevant to practitioners building agent platforms, complex automation, and multi-step job systems who need to understand how to architect for 420M+ monthly executions without latency collapse.
The key facts
14 to know420 million monthly workflow executions
Workflow size increased 12x: 2,500 to 30,000 tasks per workflow
p99 latency reduced 40%
Architecture: separated workflow metadata from task data
Asynchronous task evaluation introduced
Dynamic worker allocation and concurrency controls added
Conductor 4.0 release
420 million monthly workflow executions at Netflix
Supported workflow size increased from ~2,500 to 30,000 tasks
P99 workflow evaluation latency reduced by ~40%
Conductor 4.0 separates workflow metadata from task data
Moved evaluation to asynchronous processing
Introduces dynamic worker allocation and concurrency controls
Open-source platform (Conductor)
Go to the source
InfoQ AI/MLinfoq.com
Publisher excerpt: Netflix has reworked its Conductor workflow orchestration engine to handle larger workloads, increasing supported workflow size from about 2,500 to 30,000 tasks and reducing p99 workflow evaluation latency by about 40%. Conductor 4.0 separates workflow metadata from task data, moves evaluation to…