AI agent glossary
Durable execution
Running a long, multi-step process so its progress is saved at each step and it can resume after a crash, restart, or timeout instead of starting over.
Agents in production often run for minutes or hours and make many external calls. Servers restart, network requests fail, and deployments happen mid-task. Without protection, a failure late in the process throws away all the earlier work.
Durable execution records the result of each step as it completes. If the process dies, it restarts, replays the recorded results, and picks up where it left off. Workflow engines built for this purpose handle the bookkeeping, but a simple version can be built with a database and careful checkpointing.
It pairs naturally with idempotent tools, meaning tools that are safe to run twice, so a retried step does not, for example, send the same email again.
Learn more in the course
Related terms
Go from definitions to doing
Learn to use and build AI agents in 65 practical lessons.