Back to Blog & resources
Videos

‍Durable Execution with Dapr & Diagrid: Code-First Workflows for Cloud-Native and AI Agents

As AI agents become more autonomous, orchestrating their behavior within reliable systems is critical. Dapr Workflows, now supporting LLM-driven agents, enables structured, fault-tolerant coordination. Dapr Agents and Workflows power agentic AI with durable execution, event-driven scalability, and built-in observability. See how Dapr simplifies multi-agent automation, decision-making, and inter-agent collaboration for production-ready AI systems.

Why durable execution now

Modern distributed systems need resilient, long-running processes that survive restarts, retries, and failures. In this session, Kendall Roden shows how Dapr—an open-source, CNCF project—and Diagrid together deliver durable execution, workflow orchestration, and agent-based patterns without locking you to a single cloud, language, or runtime.

Dapr: portable building blocks for distributed apps

Dapr provides portable building-block APIsservice invocation, pub/sub, state management, durable workflows, and a conversation API—exposed via a sidecar architecture. These language-agnostic APIs abstract infrastructure, reduce plumbing code, and let developers focus on business logic while platform teams standardize on a unifying API layer across environments.

Diagrid Enterprise: operational maturity for Dapr in production

As teams move from prototype to production, Diagrid provides enterprise-grade support (hotfixes, CVE patches, upstream contributions), onboarding, and training.

  • Diagrid Dapr Distribution (D3E): enterprise-tailored Dapr with namespace-level control-plane isolation, reduced Kubernetes permissions, and integrations like Dynatrace—ideal for regulated, multi-tenant clusters.
  • Diagrid Conductor: connects your Kubernetes clusters and automates Dapr upgrades, certificate rotation, configuration reconciliation, and safe rollouts. Conductor adds real-time advisories for security, reliability, performance, and observability, plus a live app graph and per-app metrics (available on the Azure Marketplace).

Catalyst: fully managed Dapr, no sidecars to run

Diagrid Catalyst is a fully managed Dapr platform that delivers Dapr APIs as a service across Kubernetes, containers, VMs, serverless, and edge—without operating the runtime yourself. Instead of co-locating sidecars, Catalyst exposes a remote sidecar (Catalyst App ID) over the network, decoupling services from infra constraints and enabling platform independence.

Catalyst unlocks the latest capabilities—like the Dapr Workflow API—so teams can orchestrate long-running, event-driven workflows with full execution visibility and troubleshooting tools (inputs/outputs history, retries, status, and a progressive workflow execution graph).

Beyond workflows: production-grade AI agents on Dapr

Built on Dapr’s stateful workflow engine, Dapr Agents lets teams create LLM-powered agents that reason, act, and collaborate using tool calls, memory, and pub/sub. An LLM orchestrator coordinates specialized agents, routes tasks, and leverages Dapr’s automatic retries and state durability for enterprise-grade resilience. Catalyst provides message topics, a built-in key-value store (agent registry), and rich visibility into both tool-calling and orchestrator workflows.

Get started

  • Adopt Dapr’s portable APIs for resilient microservices.
  • Use Diagrid Enterprise (D3E + Conductor) for secure, observable operations on Kubernetes.
  • Choose Catalyst to run Dapr as a managed service, unlock durable workflows, and scale AI agent systems—without managing sidecars or control planes.

Try Catalyst and join the Diagrid community to accelerate durable execution, workflow automation, and AI agents—backed by enterprise support.