--- title: "Agentic workflow vs automation | Len P. van der Hof" description: An automation repeats a path. An agentic workflow holds a goal, chooses tools, and can change course. That only works with a charter, named checkpoints, and a ceiling. image: "https://lenvanderhof.com/media/generated/blog-hero-workflow-en-v2.f4af9a2904c3.wide.webp" --- [AI Systems](https://lenvanderhof.com/en/blog/category/ai-systems/) Research guide # An agentic workflow is not an automation A script repeats a path. An agentic workflow holds a goal, uses tools, and needs a ceiling. Len P. van der HofPublished 16 August 20264 min read A script repeats a path. An agent needs a ceiling. Direct answer An automation runs a fixed path. Ticket in, fields filled, Slack ping. The diagram does not invent a new diagram. An agentic workflow holds a goal, uses tools, and may change the next step when a tool returns something messy. That is useful only if someone wrote a charter, named the human checkpoints, and set an authority ceiling. Without those, you did not design a workflow. You put a junior hire with root on a looping path. ## Key takeaways - Automation = fixed path. Agentic workflow = goal plus tools plus a next step that can change. - If the path cannot surprise you, it is automation. Treat it like a script. - If the path can surprise you, it needs a charter, a packet at every handoff, and a named reviewer. - SENSE decides which workflows deserve an agent. STACK versions the commands. CHORUS names the checkpoints. This page is the split. A founder wires a model into a Zapier path and calls it an agentic workflow. The ticket still arrives. The fields still fill. Slack still pings. The only new thing is a sentence generator sitting on a rail. That is automation with a language model in one box. It is allowed. It is not an agentic workflow. ## Two machines An **automation** runs a fixed path. You can draw it. When it fails, it fails on a node you already named. A model may classify the ticket. The path does not invent a new path. An **[agentic workflow](https://lenvanderhof.com/glossary/ai-agent/)** holds a goal, uses tools, and may change the next step when a tool returns something messy. The diagram is a claim about intent, not a guarantee about the route. That is the useful part. It is also the dangerous part. The [definition page](https://lenvanderhof.com/en/blog/what-is-an-ai-agent/) is the test for the worker. This page is the test for the path the worker is allowed to walk. If the path cannot surprise you, keep calling it automation. Give it logging and an owner. Do not give it a charter you will not read. If the path can surprise you, you now have a junior operator on a looping brief. Treat it that way or unplug it. ## What has to be written down Four artifacts. If any is missing, you do not have an agentic workflow. You have a demo that can spend money. **Goal.** One sentence that survives more than one prompt. “Triage inbound” is not a goal. “Propose a next action for every inbound that matches the ICP, and stop on anything that spends, mails a customer, or changes production data” is a goal. **Charter.** The same seven fields I use for a human role in [CHORUS](https://lenvanderhof.com/frameworks/chorus/): purpose, inputs, outputs, authority ceiling, success criteria, review requirement, escalation. A worker without a charter is a guess you keep paying for. **Checkpoints.** Short workflows, not one heroic pipeline. [The Agentic Codebase](https://lenvanderhof.com/books/the-agentic-codebase/) is blunt here: turn the operation you rerun forty times a week into a versioned command, then chain those commands with named human gates. Serial where order matters. Fan-out where work can split. Fan-in where it must merge. A human gate where the next step can spend, speak as the company, or close a judgment you still own. **Packet.** When work crosses a boundary, human to agent or agent to human, six fields travel: goal, constraints, inputs, provenance, acceptance, deadline. A paste from a chat log is not a packet. Constraints die in the paste. The [human in the loop](https://lenvanderhof.com/glossary/human-in-the-loop/) is the person who can fail the output against the acceptance line. A spectator on a dashboard is not a loop. ## Which workflows deserve an agent Most paths should stay scripts. Invoicing. Provisioning. The Friday export. A model in those boxes is decoration until it starts inventing steps. Use an agentic workflow when the next step depends on a judgment you can specify but cannot fully enumerate. Qualify a lead whose form is a mess. Draft a reply that must cite a named source. Open a pull request against a written acceptance test. [SENSE](https://lenvanderhof.com/frameworks/sense/) is how you decide *where* that belongs in the week. Scan for the signal. Evaluate whether it is worth a standing system. Navigate the assumption. Simulate the cheap failure. Execute only if a Monday decision changes. [AI Agents for Startup Strategy](https://lenvanderhof.com/ai-agents-for-startup-strategy/) installs that layer. This page does not retell it. If you cannot name the Monday decision, keep the script. ## Where the other two books sit [STACK](https://lenvanderhof.com/frameworks/stack/) is what happens after you say yes. The workflow lives in the repo, not in someone’s memory of the demo. Rules in one place. Tools with contracts. [MCP](https://lenvanderhof.com/en/blog/mcp-explained-for-founders/) is how a step is allowed to touch the building. It is not the workflow. [CHORUS](https://lenvanderhof.com/frameworks/chorus/) is the coordination protocol when more than one worker shares the path. Charter, Handoff, Orchestrate, Review, Update, Sync. [The Multi-Agent Organization](https://lenvanderhof.com/books/multi-agent-organization/) is the book. The [introduction](https://lenvanderhof.com/en/blog/the-multi-agent-organization-introduction/) is that book in article form. This page is earlier: you bolted a model onto a rail and used the wrong noun. ## A twelve-minute audit Pick one path you already call agentic. 1. Draw the happy path in five boxes or fewer. 2. Circle every box that can choose a tool the diagram does not name. 3. Write the goal in one sentence. 4. Write the one action that box is forbidden to take. 5. Name the human who can stop that action. 6. Point at the file where the command is versioned. Not the chat. The file. If step 2 is empty, you have automation. Good. Stop saying agentic. If step 2 has circles and steps 3 to 6 are blank, you have an unchartered operator on a loop. Fill the lines or take the tools away. The vocabulary is the control. ## Terminology - [AI agent](https://lenvanderhof.com/glossary/ai-agent/) - [Human in the loop](https://lenvanderhof.com/glossary/human-in-the-loop/) - [CHORUS](https://lenvanderhof.com/glossary/chorus/) ## Sources 1. [AI agent (glossary)](https://lenvanderhof.com/glossary/ai-agent/) 2. [SENSE framework](https://lenvanderhof.com/frameworks/sense/) 3. [STACK framework](https://lenvanderhof.com/frameworks/stack/) 4. [CHORUS framework](https://lenvanderhof.com/frameworks/chorus/) 5. [AI Agents for Startup Strategy](https://lenvanderhof.com/ai-agents-for-startup-strategy/) 6. [The Agentic Codebase](https://lenvanderhof.com/books/the-agentic-codebase/) 7. [The Multi-Agent Organization](https://lenvanderhof.com/books/multi-agent-organization/) ## Related essays - [What is an AI agent?](https://lenvanderhof.com/en/blog/what-is-an-ai-agent/) - [MCP explained for founders](https://lenvanderhof.com/en/blog/mcp-explained-for-founders/) - [AI Agents for Startup Strategy: introduction](https://lenvanderhof.com/en/blog/ai-agents-for-startup-strategy-introduction/) - [The Agentic Codebase: introduction](https://lenvanderhof.com/en/blog/the-agentic-codebase-introduction/) - [The Multi-Agent Organization: introduction](https://lenvanderhof.com/en/blog/the-multi-agent-organization-introduction/) - [Why multi-agent systems fail](https://lenvanderhof.com/en/blog/why-multi-agent-systems-fail/) - [A SENSE intelligence layer in 90 days](https://lenvanderhof.com/en/blog/sense-90-day-intelligence-layer/) - [Version the agent OS](https://lenvanderhof.com/en/blog/stack-version-the-agent-os/) - [LLM integration is an operating-design problem](https://lenvanderhof.com/en/blog/llm-integration-is-operating-design/) - [A go-to-market system that survives the first ten customers](https://lenvanderhof.com/en/blog/gtm-that-survives-ten-customers/) - [AI agent or chatbot: three tests before you mix the words](https://lenvanderhof.com/en/blog/ai-agent-or-chatbot/) - [MCP meaning: Model Context Protocol, not a product](https://lenvanderhof.com/en/blog/mcp-meaning/) - [What is STACK? Five layers, not a vendor diagram](https://lenvanderhof.com/en/blog/what-is-stack/) - [What is CHORUS? Coordination, not a choir](https://lenvanderhof.com/en/blog/what-is-chorus/) ## Further reading - [What is an AI agent?](https://lenvanderhof.com/en/blog/what-is-an-ai-agent/) - [MCP explained for founders](https://lenvanderhof.com/en/blog/mcp-explained-for-founders/) - [SENSE](https://lenvanderhof.com/frameworks/sense/) - [STACK](https://lenvanderhof.com/frameworks/stack/) - [CHORUS](https://lenvanderhof.com/frameworks/chorus/) - [AI Agents for Startup Strategy](https://lenvanderhof.com/ai-agents-for-startup-strategy/) - [The Agentic Codebase](https://lenvanderhof.com/books/the-agentic-codebase/) - [The Multi-Agent Organization](https://lenvanderhof.com/books/multi-agent-organization/) About the author ## [Len P. van der Hof](https://lenvanderhof.com/en/authors/len-p-van-der-hof/) Entrepreneur, AI Innovator and Venture Builder Len P. van der Hof builds practical AI systems, digital ventures and evidence-informed tools for founders. ```json { "@context": "https://schema.org", "@graph": [ { "@type": "Person", "@id": "https://lenvanderhof.com/#person", "name": "Len P. van der Hof", "alternateName": [ "Len van der Hof", "L.P. van der Hof", "Leendert Pieter van der Hof" ], "honorificSuffix": "MSc", "url": "https://lenvanderhof.com/", "image": [ "https://lenvanderhof.com/photos/len-portrait-1.jpg", "https://lenvanderhof.com/photos/len-portrait-2.jpg", "https://lenvanderhof.com/photos/len-portrait-3.jpg", "https://lenvanderhof.com/photos/len-portrait-4.jpg", "https://lenvanderhof.com/photos/len-speaking.jpg", "https://lenvanderhof.com/photos/len-hero.jpg" ], "jobTitle": "Entrepreneur, AI Innovator and Venture Builder", "description": "Len P. van der Hof is a Dutch entrepreneur and AI innovator building ReasonKit, MindSesh, Undominated.ai, the fiction imprint LPH98.lifestyle, and technology ventures through LPH98.ventures.", "address": { "@type": "PostalAddress", "addressLocality": "Zwijndrecht", "addressCountry": "NL" }, "alumniOf": { "@type": "CollegeOrUniversity", "name": "Rotterdam School of Management, Erasmus University" }, "knowsAbout": [ "Artificial intelligence", "AI agents", "Agentic AI systems", "LLM routing", "SEO", "Generative engine optimization", "Answer engine optimization", "Venture building", "Founder performance", "Founder psychology", "Evidence-based decision-making" ], "sameAs": [ "https://www.linkedin.com/in/lenvanderhof/", "https://x.com/LenvanderHof", "https://www.youtube.com/channel/UCTG20buKqYYbitqqf7l3zJA", "https://www.instagram.com/Lenvanderhof/", "https://www.threads.com/@lenvanderhof", "https://github.com/Lenvanderhof", "https://huggingface.co/LPH98", "https://www.npmjs.com/~lenvanderhof", "https://www.goodreads.com/author/show/70983905.Len_P_van_der_Hof", "https://www.amazon.com/author/lenvanderhof", "https://www.bol.com/nl/nl/b/len-p-van-der-hof-msc/609879394/", "https://bsky.app/profile/lenvanderhof.com", "https://mastodon.social/@Lenvanderhof", "https://crates.io/users/Lenvanderhof", "https://cursor.com/@Lenvanderhof", "https://medium.com/@Lenvanderhof", "https://gitlab.com/Lenvanderhof", "https://hub.docker.com/u/lenvanderhof/", "https://dev.to/lenvanderhof", "https://www.facebook.com/Lenvanderhof", "https://soundcloud.com/Lenvanderhof" ], "affiliation": [ { "@id": "https://lenvanderhof.com/#publisher" }, { "@id": "https://lenvanderhof.com/#mindsesh" }, { "@id": "https://lenvanderhof.com/#lifestyle" } ] }, { "@type": "WebSite", "@id": "https://lenvanderhof.com/#website", "url": "https://lenvanderhof.com/", "name": "Len P. van der Hof", "description": "Len P. van der Hof is a Dutch entrepreneur and AI innovator building ReasonKit, MindSesh, Undominated.ai, the fiction imprint LPH98.lifestyle, and technology ventures through LPH98.ventures.", "inLanguage": [ "en", "nl" ], "publisher": { "@id": "https://lenvanderhof.com/#person" } }, { "@type": "Organization", "@id": "https://lenvanderhof.com/#publisher", "name": "LPH98.ventures", "url": "https://lph98.ventures", "founder": { "@id": "https://lenvanderhof.com/#person" } }, { "@type": "Organization", "@id": "https://lenvanderhof.com/#mindsesh", "name": "MindSesh", "url": "https://mindsesh.net", "founder": { "@id": "https://lenvanderhof.com/#person" } }, { "@type": "SoftwareApplication", "@id": "https://lenvanderhof.com/#reasonkit", "name": "ReasonKit", "url": "https://reasonkit.sh", "creator": { "@id": "https://lenvanderhof.com/#person" } }, { "@type": "SoftwareApplication", "@id": "https://lenvanderhof.com/#undominated", "name": "Undominated.ai", "url": "https://undominated.ai", "creator": { "@id": "https://lenvanderhof.com/#person" } }, { "@type": "Organization", "@id": "https://lenvanderhof.com/#lifestyle", "name": "LPH98.lifestyle", "url": "https://lph98.lifestyle", "founder": { "@id": "https://lenvanderhof.com/#person" } }, { "@type": "ImageObject", "@id": "https://lenvanderhof.com/en/blog/agentic-workflow-vs-automation/#primaryimage", "url": "https://lenvanderhof.com/media/generated/blog-hero-workflow-en-v2.f4af9a2904c3.wide.webp", "contentUrl": "https://lenvanderhof.com/media/generated/blog-hero-workflow-en-v2.f4af9a2904c3.wide.webp", "representativeOfPage": true }, { "@type": "BreadcrumbList", "@id": "https://lenvanderhof.com/en/blog/agentic-workflow-vs-automation/#breadcrumb", "itemListElement": [ { "@type": "ListItem", "position": 1, "name": "Home", "item": "https://lenvanderhof.com/" }, { "@type": "ListItem", "position": 2, "name": "Blog", "item": "https://lenvanderhof.com/en/blog/" }, { "@type": "ListItem", "position": 3, "name": "AI Systems", "item": "https://lenvanderhof.com/en/blog/category/ai-systems/" }, { "@type": "ListItem", "position": 4, "name": "An agentic workflow is not an automation", "item": "https://lenvanderhof.com/en/blog/agentic-workflow-vs-automation/" } ] }, { "@type": "WebPage", "@id": "https://lenvanderhof.com/en/blog/agentic-workflow-vs-automation/#webpage", "url": "https://lenvanderhof.com/en/blog/agentic-workflow-vs-automation/", "name": "An agentic workflow is not an automation", "description": "An automation repeats a path. An agentic workflow holds a goal, chooses tools, and can change course. That only works with a charter, named checkpoints, and a ceiling.", "isPartOf": { "@id": "https://lenvanderhof.com/#website" }, "primaryImageOfPage": { "@id": "https://lenvanderhof.com/en/blog/agentic-workflow-vs-automation/#primaryimage" }, "breadcrumb": { "@id": "https://lenvanderhof.com/en/blog/agentic-workflow-vs-automation/#breadcrumb" }, "inLanguage": "en-GB" }, { "@type": "BlogPosting", "@id": "https://lenvanderhof.com/en/blog/agentic-workflow-vs-automation/#article", "mainEntityOfPage": { "@id": "https://lenvanderhof.com/en/blog/agentic-workflow-vs-automation/#webpage" }, "headline": "An agentic workflow is not an automation", "description": "An automation repeats a path. An agentic workflow holds a goal, chooses tools, and can change course. That only works with a charter, named checkpoints, and a ceiling.", "datePublished": "2026-08-16T20:30:00.000Z", "author": { "@id": "https://lenvanderhof.com/#person" }, "publisher": { "@id": "https://lenvanderhof.com/#person" }, "image": [ "https://lenvanderhof.com/media/generated/blog-hero-workflow-en-v2.f4af9a2904c3.square.webp", "https://lenvanderhof.com/media/generated/blog-hero-workflow-en-v2.f4af9a2904c3.landscape.webp", "https://lenvanderhof.com/media/generated/blog-hero-workflow-en-v2.f4af9a2904c3.wide.webp" ], "articleSection": "AI Systems", "inLanguage": "en-GB" } ] } ```