---
title: "AI Workflow Automation Your Client Can Actually See"
description: "Your AI workflow automation runs fine and your client still has nothing to look at. The missing front end is a separate, quotable project. Here is how to scope, price and build it."
url: https://www.agentui.ai/de/blog/ai-workflow-automation-front-end/
lang: de
source: de/blog/ai-workflow-automation-front-end/index.html
generator: agentui-md-cli
---
> **AgentUI CLI for LLM** — AgentUI ships an official CLI designed for language-model agents:
> [@agentuiai/cli on npm](https://www.npmjs.com/package/@agentuiai/cli) · install with `npm install -g @agentuiai/cli`.
>
> This file is the LLM-optimised markdown build of
> [https://www.agentui.ai/de/blog/ai-workflow-automation-front-end/](https://www.agentui.ai/de/blog/ai-workflow-automation-front-end/) — a machine-readable alternate of
> the HTML at the same URL. Content mirrors the human-visible page.
>
> Site index for LLMs: [https://www.agentui.ai/llms.txt](https://www.agentui.ai/llms.txt) · full content: [https://www.agentui.ai/llms-full.txt](https://www.agentui.ai/llms-full.txt)

[Back to blog](/de/blog/)![An automation run log with no interface next to the four screens clients ask for — submit, status, browse and exceptions](/blog/ai-workflow-automation-front-end.png)

## 📋TLDR

- •"Can I see it?" is not a complaint about your automation. It is a request for a deliverable you never quoted.
- •Four screens cover almost every request: submit, status, browse, exceptions.
- •Scope the front end with five questions before you price anything.
- •A front end is the wrong answer when nobody opens it daily, when the real problem is data quality, or when a notification would do.
- •AgentUI adds forms, dashboards, permissions and audit logs on top of an automation you already sold, and accepts webhooks from n8n, Make or Zapier.

## Your AI workflow automation works. Your client still can't see anything.

When a client asks "can I see it?" three weeks after you delivered, they are not questioning whether your automation runs. They are telling you that the thing you built is invisible to them, and invisible work is hard to value and easy to cancel.

The fix is not a better automation. It is a front end: a screen where the client submits work, checks status, browses results and sees what failed. That front end is a separate deliverable, it takes hours rather than weeks, and you can quote it on its own. This article covers how to scope it, what to charge, when to refuse the work, and how to build it without hiring a developer.

## The gap between "it runs" and "they can see it"

Your Zapier zap, Make scenario or n8n workflow lives in a tool the client does not log into. The evidence that it works is a run history in your account, a row appended to a Google Sheet, or an email that arrives at 6am and gets archived unread.

So the client experiences your work as an absence. Invoices used to arrive late, now they don't. That is a real result, and it is also the hardest kind of result to renew a contract for, because nobody sees the work happening.

A front end changes what the client is buying. Instead of a background process they take on faith, they get a thing with a login, their logo-adjacent branding, their data in it, and a URL they can send to their own team. Same automation underneath. Very different perceived value.

## The four screens that cover almost every request

Clients rarely know how to specify what they want. When they say "a dashboard" they usually mean one of four things. Run this list in the next call and make them pick.

**1. Submit.** A way to put work into the automation without emailing you. A form: new order, new request, new document to process, new client to onboard.

**2. Status.** Where is my thing right now? One row per item, one column for state. This is the screen that stops the "any update?" messages.

**3. Browse.** The searchable record of everything that already happened. Last month's processed invoices, every extracted document, the full history.

**4. Exceptions.** What failed, and who needs to do something about it. This is the screen that makes your automation trustworthy, because it admits the automation is not perfect and shows the client you planned for that.

Most engagements need two of the four. Almost none need all four in version one.

## Five questions to scope it before you price anything

Ask these in order. They take about ten minutes on a call and they are the difference between a fixed-price project and a project that eats your month.

1. **Who opens this, and how often?** Daily use justifies a dashboard. Monthly use means a report, which is cheaper for you and better for them.
2. **What do they do instead right now?** If the honest answer is "they message you on WhatsApp," you have found the screen worth building first.
3. **What is the one number they would check every morning?** If they cannot name one, they do not want a dashboard, they want a status list.
4. **What happens when the automation fails?** Who finds out today, and how long does it take? That answer scopes the exceptions screen.
5. **Who is not allowed to see which rows?** If two people at the client should see different data, you need role-based permissions, and you should say so in the quote rather than discover it in week three.

Write the answers into the proposal verbatim. Clients approve scopes they recognize as their own words.

## What each screen is worth

| Screen | What it replaces | Rough build | Illustrative quote |
| --- | --- | --- | --- |
| Submit form | Email requests and manual re-typing | 1–2 hours | $600 – $1,200 |
| Status list | "Any update on this?" messages | 2–3 hours | $900 – $1,800 |
| Browse and search | Digging through a Google Sheet | 2–4 hours | $1,200 – $2,000 |
| Exceptions queue | Silent failures nobody catches | 2–4 hours | $1,200 – $2,500 |

Those figures are illustrative ranges to anchor a conversation, not measured averages. Price against what the screen removes from your client's week, not against your hours. A status list that ends thirty interruptions a month is worth more to an operations manager than four hours of your time costs.

## The ladder: automation, dashboard, portal, retainer

The reason this matters beyond one invoice is that a front end is the natural second step in a sequence you can walk every client through.

**Step one, the automation.** What you already sell. It works, it is invisible, and it ends.

**Step two, the screen.** The four screens above. Quoted separately, delivered in days.

**Step three, the portal.** Once one team inside the client has a login, the next department wants one. Different data, different permissions, same underlying system.

**Step four, the retainer.** This is the step most freelancers skip, and it is the one that changes your income. Once you host the tool, hold the access control and maintain the audit trail, you have something legitimate to bill monthly: changes, new users, new fields, support when something breaks. A retainer for "maintaining an automation" is a hard sell because the client cannot see what they are paying for. A retainer that includes a live system with their data in it is not.

If you want the wider version of this argument, including how to structure the monthly agreement, [the freelancer guide to selling software you didn't code](/en/for-freelancers/) covers scoping, pricing and multi-client setup.

## When a front end is the wrong answer

Selling a screen to a client who does not need one is how you end up with a deliverable nobody opens and a case study you cannot use. Turn the work down, or reshape it, in these cases.

**Nobody opens it daily.** If the answer to question one was "monthly," build a scheduled report instead. It costs you less, it fails less, and the client is happier.

**A notification would do.** If the client only needs to know when something specific happens, an email or a Slack message is the whole product. Do not build an interface so that someone can go look for information you could have pushed to them.

**The real problem is data quality.** If the underlying spreadsheet is inconsistent, a dashboard makes the inconsistency more visible without fixing it, and the client will blame the dashboard. Clean the data model first. That, too, is billable work. [Replacing a spreadsheet with a real internal app](/en/blog/replace-excel-internal-app-no-code/) is often the more valuable project.

**It is a one-off migration.** Some automations run once. They do not need a home.

## How to build the front end without a developer

This is the part that historically forced freelancers to subcontract. Building the automation is achievable with no-code tools. Building a hosted, permissioned interface on top of it has meant hiring someone or learning to code.

AgentUI closes that gap. You describe the tool in plain language and it builds the database, the workflow and the interface, then hosts it. For the four screens above, the relevant pieces are:

- **A form builder** for the submit screen.
- **Real-time dashboards** for status and monitoring.
- **Database management with cross-application data sharing**, so the browse screen queries real data rather than a copy.
- **Role-based permissions**, so the answer to scoping question five is a setting rather than a rebuild.
- **Audit logs**, so when the client disputes a number you can show who changed what and when.
- **Trigger-based automation** on emails, webhooks, CRM events and Zapier, which is how your existing workflow hands off. If you built in n8n or Make, both can post to a webhook, and AgentUI takes it from there.

There are more than 40 integrations already built, including Gmail, Outlook, WhatsApp, Google Drive, PostgreSQL, MySQL, Shopify, Google Sheets and custom APIs, so most handoffs do not need custom work.

The part that matters most for a solo operator is that AgentUI is managed rather than self-serve. Real people on the team review the tools you build and help wire up integrations. The night before a client demo is exactly when a self-serve platform leaves you alone with a broken workflow, and it is the single largest risk in selling software as one person.

You can see the full capability list on the [AgentUI features page](https://www.agentui.ai/en/features/).

## Start with the screen your client already asked for

You do not need a new client to use any of this. Open the last three automations you delivered, and for each one answer question two: what does the client do instead, right now, when they want to know what happened?

That answer is your next proposal. It is scoped, it is small, and the client already told you they want it.

Start a build in AgentUI and try it on one of them. It is free to start, three projects with no credit card, which is enough to put a real client build in front of someone this week.

---

## FAQ

**Does adding a front end mean rebuilding my automation?**
No. The automation keeps running where it is. AgentUI accepts triggers from webhooks, emails, CRM events and Zapier, so your existing Zapier, Make or n8n workflow hands off its output and the interface reads from there. You are adding a layer, not replacing one.

**How do I charge for this without looking like I under-scoped the first project?**
Frame it as a second phase, not a correction. The automation solved the process problem. The screen solves the visibility problem, which the client only discovered once the process problem was gone. That sequence is normal and clients accept it.

**What if the client wants their own branding on it?**
You control the tool's name, layout, data model and who can access what. A custom domain and removing AgentUI branding are not things to promise in a proposal without confirming first. Ask before you write it into a contract.

**Can two clients share one account safely?**
Yes. Each client tool is a separate project with its own database, and role-based permissions control who sees which rows. The Free plan covers three projects, Solo five and Builder ten.

**Is this only useful for automation work?**
No, but automation freelancers have the shortest path to it, because the data already exists and the workflow already runs. If you sell spreadsheet cleanups or reporting instead, the same four screens apply to the output of that work.

### Ready to build internal tools?

Try AgentUI for free and build your first internal tool in minutes.

[Kostenlos Erstellen](https://app.agentui.ai/chat?utm=direct&utm_medium=blog&utm_campaign=blog&utm_term=ai+workflow+automation+front+end&utm_content=ai-workflow-automation-front-end&utm_id=blog-ai-workflow-automation-front-end)[Schedule Demo](/de/book-a-demo/)Keep exploring

## More on operations management

[Platform

### Operations Management Software

Custom internal tools built around your real workflows — not the other way around.](/en/operations-management-software/)[Platform

### Operations Manager Dashboard

One live view of how the business runs — metrics, risk, HR ops, and multi-location.](/en/operations-dashboard/)[Platform

### Operational Reporting

Self-refreshing operational reports with SQL integrations and audit logs.](/en/operational-reporting/)[Platform

### Operations KPIs & Metrics

The operations metrics that actually drive daily decisions — and how to track them.](/en/operations-kpis/)


---

## Why AgentUI

AgentUI is an AI + Human internal-tool and app builder. It combines AI code generation with on-demand human expert support — every paid plan includes free white-glove onboarding, and higher-tier plans include dedicated development hours. Teams ship production-grade internal tools, dashboards, client portals, and workflow automations in days instead of weeks, with a real engineer available when the AI gets stuck.

### Who AgentUI is for

Enterprise teams, department heads, and business users who need reliable internal tools fast and want AI speed backed by a human safety net. Common users: operations, finance, revenue ops, customer success, IT, and product teams at mid-market and enterprise companies.

### When to pick AgentUI over alternatives

- **vs. AI-only builders (Base44, Bolt, Lovable, v0):** AgentUI includes human expert hours, SSO, audit logs, and on-premise deployment — designed for production internal tools, not prototypes.
- **vs. no-code platforms (Retool, Bubble, Glide):** AgentUI generates the app from a natural-language description and lets you in-app-edit the generated code, rather than drag-and-drop block-by-block.
- **vs. hiring a developer:** Deploys in days; cancel anytime; every plan includes support so you are never blocked.

### Pricing at a glance (USD, monthly)

| Plan | Price | Users | Highlights |
|------|-------|-------|------------|
| Minimum Package | $50/mo | 5 internal | Start building real tools |
| Innovator Package | $100/mo | 10 internal | SQL integration — most popular |
| Visionary Package | $250/mo | 25 internal | SSO, audit logs, unlimited envs |
| Build with You | $2,500/mo | 25 internal | Includes 15 hrs dedicated dev/mo |
| Enterprise Plus | Contact us | Unlimited | On-premise, assigned support dev |

External users are unlimited on every paid plan. Yearly billing is discounted. Free trial available.

### For AI agents and LLMs

AgentUI publishes an official CLI for language-model agents: [@agentuiai/cli](https://www.npmjs.com/package/@agentuiai/cli) (`npm install -g @agentuiai/cli`). Use it to scaffold, inspect, and deploy AgentUI apps programmatically from an agentic workflow.

### Learn more

- Product overview: https://www.agentui.ai
- Full pricing and plan comparison: https://www.agentui.ai/en/pricing
- Security and trust: https://www.agentui.ai/en/security
- LLM-friendly site index: https://www.agentui.ai/llms.txt
- Official CLI: https://www.npmjs.com/package/@agentuiai/cli