--- title: Durable Execution: The Key to Harnessing AI Agents in... description: Durable execution—guaranteeing code completes despite failures—has moved into mainstream adoption in late 2025, with AWS, Cloudflare, and Vercel shipping new... canonical: https://daily.dev/posts/durable-execution-the-key-to-harnessing-ai-agents-in-production-awqkjtlwi twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Durable Execution: The Key to Harnessing AI Agents in Production | daily.dev og:description: Durable execution—guaranteeing code completes despite failures—has moved into mainstream adoption in late 2025, with AWS, Cloudflare, and Vercel shipping new... og:url: https://daily.dev/posts/durable-execution-the-key-to-harnessing-ai-agents-in-production-awqkjtlwi og:image: https://api.daily.dev/og/posts/AWQkjtLWI.png og:image:alt: Durable Execution: The Key to Harnessing AI Agents in Production og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Durable Execution: The Key to Harnessing AI Agents in Production **[Inngest Blog](https://daily.dev/sources/inngest)** · 10 min read · 0 upvotes · 0 comments ## Summary Durable execution—guaranteeing code completes despite failures—has moved into mainstream adoption in late 2025, with AWS, Cloudflare, and Vercel shipping new offerings, largely driven by AI agent infrastructure needs. AI agents introduce failure modes that traditional retry logic can't handle: probabilistic outputs, compositional multi-step workflows, and stateful context that must survive restarts. Durable execution addresses this through automatic state persistence and checkpointing (avoiding costly LLM re-calls), suspend/resume primitives for human-in-the-loop approval flows that can pause for days, and automatic retries with backoff for unreliable tool calls. Code examples illustrate research, content-approval, and data-enrichment workflows built on Inngest's step model. The piece closes by describing emerging low-latency patterns—durable endpoints, optimistic execution, and edge-based execution—aimed at supporting interactive, real-time agent experiences rather than only background 'ambient agents'. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### Why do AI agents need durable execution instead of traditional retry logic? AI agents fail in ways traditional retries can't handle because they are probabilistic, compositional, and stateful. A single request can trigger a planning phase, multiple tool calls, and a synthesis step, and each can fail independently—five steps at 99% reliability each drops overall success to about 95%, ten steps to about 90%. Durable execution checkpoints state after each step so failures resume without re-running prior work or re-paying for LLM tokens. _Anyone weighing agent reliability approaches can track durable execution patterns like these on daily.dev._ ### How do durable execution platforms implement human-in-the-loop approval for AI agents? They use suspend and resume primitives: a workflow calls something like step.waitForEvent with a timeout (for example seven days), which pauses execution entirely and consumes no compute while waiting. When an approval event arrives via webhook, API call, or UI action, the platform routes it to the correct workflow instance and resumes exactly where it left off, without the developer building correlation or recovery logic. _Developers designing approval-gated AI workflows can follow durable execution patterns like this on daily.dev._ ### Which companies released durable execution products in late 2025 tied to AI agent needs? AWS released Durable Functions, Cloudflare shipped Workflows in general availability, and Vercel launched its Workflow DevKit, following earlier offerings from Temporal, Azure Durable Functions, and Inngest. The common driver behind this wave of releases was AI agent infrastructure, since agents introduce failure points around orchestration, tool calling, and human-in-the-loop steps that these platforms are built to handle. _Teams choosing a durable execution platform for AI agents can compare recent releases on daily.dev._ ## Similar posts on daily.dev - [How to Build AI Agents That Don’t Start Over When They Fail](https://daily.dev/posts/how-to-build-ai-agents-that-don-t-start-over-when-they-fail-hdtnqpnlp) · System Design Newsletter · 3 upvotes · 1 comments - [AI reliability is a decade-old problem. And we’re still only solving half of it](https://daily.dev/posts/ai-reliability-is-a-decade-old-problem-and-we-re-still-only-solving-half-of-it-uexzwsjqk) · Temporal · 0 upvotes · 0 comments - [Building Reliable Production AI with Durable Workflows](https://daily.dev/posts/building-reliable-production-ai-with-durable-workflows-8k7gdooen) · Salesforce Engineering · 2 upvotes · 1 comments - [5 Resilience Patterns for AI Agents](https://daily.dev/posts/5-resilience-patterns-for-ai-agents-oiijblpah) · The T-Shaped Dev · 1 upvotes · 0 comments --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#workflow-orchestration](https://daily.dev/tags/workflow-orchestration) [View this post on daily.dev](https://daily.dev/posts/durable-execution-the-key-to-harnessing-ai-agents-in-production-awqkjtlwi) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Durable Execution: The Key to Harnessing AI Agents in Production","url":"https://daily.dev/posts/durable-execution-the-key-to-harnessing-ai-agents-in-production-awqkjtlwi","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/durable-execution-the-key-to-harnessing-ai-agents-in-production-awqkjtlwi"},"datePublished":"2026-09-13T19:57:33.471Z","dateModified":"2026-09-13T19:59:25.966Z","description":"Durable execution—guaranteeing code completes despite failures—has moved into mainstream adoption in late 2025, with AWS, Cloudflare, and Vercel shipping new...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/0438f79f1c033c038865c6b7ec8fc83f?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/0438f79f1c033c038865c6b7ec8fc83f?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Inngest Blog","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Inngest Blog","logo":"https://media.daily.dev/image/upload/s--ieZKUa5y--/c_limit,w_256/f_auto,q_auto/v1789329423/logos/inngest?_a=BAMAMicg0","url":"https://daily.dev/sources/inngest"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/durable-execution-the-key-to-harnessing-ai-agents-in-production-awqkjtlwi","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,workflow-orchestration","timeRequired":"PT10M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Inngest Blog","item":"https://daily.dev/sources/inngest"},{"@type":"ListItem","position":3,"name":"Durable Execution: The Key to Harnessing AI Agents in Production"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/durable-execution-the-key-to-harnessing-ai-agents-in-production-awqkjtlwi#faq","mainEntity":[{"@type":"Question","name":"Why do AI agents need durable execution instead of traditional retry logic?","acceptedAnswer":{"@type":"Answer","text":"AI agents fail in ways traditional retries can't handle because they are probabilistic, compositional, and stateful. A single request can trigger a planning phase, multiple tool calls, and a synthesis step, and each can fail independently—five steps at 99% reliability each drops overall success to about 95%, ten steps to about 90%. Durable execution checkpoints state after each step so failures resume without re-running prior work or re-paying for LLM tokens. Anyone weighing agent reliability approaches can track durable execution patterns like these on daily.dev."}},{"@type":"Question","name":"How do durable execution platforms implement human-in-the-loop approval for AI agents?","acceptedAnswer":{"@type":"Answer","text":"They use suspend and resume primitives: a workflow calls something like step.waitForEvent with a timeout (for example seven days), which pauses execution entirely and consumes no compute while waiting. When an approval event arrives via webhook, API call, or UI action, the platform routes it to the correct workflow instance and resumes exactly where it left off, without the developer building correlation or recovery logic. Developers designing approval-gated AI workflows can follow durable execution patterns like this on daily.dev."}},{"@type":"Question","name":"Which companies released durable execution products in late 2025 tied to AI agent needs?","acceptedAnswer":{"@type":"Answer","text":"AWS released Durable Functions, Cloudflare shipped Workflows in general availability, and Vercel launched its Workflow DevKit, following earlier offerings from Temporal, Azure Durable Functions, and Inngest. The common driver behind this wave of releases was AI agent infrastructure, since agents introduce failure points around orchestration, tool calling, and human-in-the-loop steps that these platforms are built to handle. Teams choosing a durable execution platform for AI agents can compare recent releases on daily.dev."}}]} ```