edit

Dashboards and Visualizations

Metrics, dashboards, and visualizations are native to PipelineAI.

We believe that you should have full insight into everything deployed to production.

PipelineAI provides the usual dashboards for real-time system metrics including memory usage, disk I/O and network throughput, CPU and GPU utilization of your model training and deployment activities.

In addition, PipelineAI provides dashboards for real-time prediction metrics including accuracy, latency, and throughput of your models in production.

Example Dashboards

Stabilize the Cluster

Unhealthy or latent model servers may open a circuit, respond with a suitable fallback, and allow the cluster to stabilize.

Stabilize the Cluster

Optimize Performance

Large batch sizes provide higher throughput at the expense of latency. PipelineAI dynamically configures the system to find the proper balance.

Optimize Performance

Control Latency with Timeouts

High latency may lead to unhealthy model servers if left unbounded. All PipelineAI service calls are bound with timeouts.

Monitor and Alert

High resource utilization - beyond container and physical node limits - will certainly degrade performance. PipelineAI monitors all system resources.

Scale Dynamically

All PipelineAI services support auto-scaling across federated cloud and on-premise environments.


Try our Free Community Edition!

Click HERE to get started with our basic Community Edition

Register for PipelineAI Advanced Edition


Take a Quick, Fun Survey!


Help Us Improve

Option 1: Edit the documentation directly.

Option 2: Create a GitHub issue.