In the news
Durable execution in Pydantic AI agents with AWS Lambda Durability
Pydantic · Laís Carvalho · Published · 3 min read
In 30 seconds
- What happened
- Pydantic AI Harness now includes AWSLambdaDurability, a capability that checkpoints every model request and tool call so Lambda handlers can resume after timeouts or retries without duplicating work.
- Why it matters
- Matters for engineers building AI agents on AWS Lambda that run longer than the invocation timeout or need to survive retries without re-executing expensive operations.
- Watch out
- Changing agent shape, tool count, or model breaks in-flight executions. Tools run sequentially inside durable handlers, not in parallel. Requires specific IAM permissions and Lambda durable configuration.
- agent
- mcp
The patterns behind this
- Durable Execution & Checkpointing
- MCP Gateway (Tool Federation & Governance)
- Speculative & Parallel Tool Execution
Each one covers how the technique works, when it earns its cost, and where it breaks.
The Agent Architect
One pattern, one tradeoff, one production failure story. A short weekly briefing for people building agentic systems.
Weekly email, one-click unsubscribe. We only use your address to send the briefing.