# Pydantic AI + Dapr

Make your Pydantic AI agents production-ready. These tutorials show you how to add fault tolerance using Dapr.

- [Durable Workflows](https://docs.diagrid.io/getting-started/quickstarts/ai-agents?agentframework=pydantic-ai#1-log-in-to-catalyst) — Make Pydantic AI agents immune to failures with automatic recovery and replay.
- [Access Policies](https://docs.diagrid.io/develop/agents/pydantic-ai/pydantic-ai-access-policies) — Secure agent-to-agent communication with zero-trust access policies.

---

## Why Dapr for Pydantic AI?

| Challenge | Dapr Solution |
|-----------|---------------|
| Agents tightly coupled, hard to scale | **Pub/sub** decouples communication |
| Chat memory lost on restart | **State store** persists agent memory across restarts |
| Custom infrastructure for each deployment | **Portable components** work anywhere |
| No visibility into agent communication | **Built-in observability** traces everything |