Diagrid

Dapr 1.19 Cuts AI Workflow Latency and Simplifies the Control Plane

Dapr 1.19 Cuts AI Workflow Latency and Simplifies the Control Plane

Diagrid has released Dapr 1.19, bringing delta-history transmission and a WorkflowsFastPath mode that cuts median workflow step latency from 553ms to 153ms. The release also consolidates two overlapping control-plane services into one, reducing operational complexity for teams running durable AI agent workloads. For government and regulated-sector operators, the combination of immutable workflow history and CNCF governance makes Dapr an increasingly credible infrastructure choice.

Dapr 1.19 Cuts AI Workflow Latency and Simplifies the Control Plane Read More »

Diagrid Catalyst Brings Durable Execution to AI Agents on Kubernetes

Diagrid Catalyst Brings Durable Execution to AI Agents on Kubernetes

Diagrid has extended its Catalyst platform to bring durable execution to Goose and 15+ agent frameworks, ensuring long-running AI workflows survive crashes without restarting. Built on the CNCF-graduated Dapr runtime, Catalyst runs portably across AWS, Azure, Google Cloud, and on-premises Kubernetes. For enterprises moving agents into production, this is the reliability layer the market has been missing.

Diagrid Catalyst Brings Durable Execution to AI Agents on Kubernetes Read More »