Step 1 · We map the workflowHow many steps does this job take today?
Workflow automationLet the repeating steprun in the background.
Invoicing, data entry, sending reports, notifications and checks run automatically. The process picks the tool: where the data has to stay, how many steps the workflow has and which systems it connects to make the choice.
The tools we use
- n8n
Make
Zapier
Airtable
Supabase
GitHub
Docker
How it is built
From the step done by handto the step running in the background.
First we work out the steps and the exceptions. Only then is the tool chosen, and the reasoning written down. Once the workflow is built, run health starts being watched.
Step 2 · The tool is chosen and the workflow builtThe reasoning is written, the workflow documented.
Step 3 · It is watched and growsRun health becomes a number.
What it can do
It takes on the stepand leaves the decision to you.
Six modules build the same skeleton: trigger, log, error queue and approval point. Because the rule is written, the outcome does not change with who built the workflow.
A repeating step, a defined workflow
Every workflow is documented with its trigger, its steps and its failure behaviour. If the person who built it leaves, whoever reads the document carries it on.
The process picks the tool
The choice looks at three measures: whether the data may leave your systems, how many steps the workflow has, and whether the systems it connects to have ready integrations. All three are measured in discovery.
Every run is logged
When the workflow ran, which record it ran on and what the outcome was are written to the audit log. Every past run is traced from that same log.
Failures land in a queue
A failed run is reported to its owner, its reason is written down and it is retried. If the retry also fails, the workflow stops and you hear about it.
Human approval can be a step
An approval step sits in the middle of the workflow: actions that are hard to undo, such as sending, paying or deleting, wait for a person.
The health of the workflow is visible
Run success rate, average duration and the number of waiting errors reach you every month. When a workflow starts slowing down, you see it in the numbers.
Choosing the tool
We do not pick the tool first,we pick it after understanding the process.
n8n runs on your own server and keeps the data in; Make is fast for visual flows; Zapier stands out for the number of ready integrations. Which one gets chosen is not a brand decision but what those three properties mean for your workflow.
- The tool is chosen after the process
- Every run is logged
- Failures queue and get reported
Frequently asked
Workflow automationFrequently asked questions
It depends on the process. If the data has to stay on your own server, n8n; if it is a visual, mid-complexity flow, Make; if it is a simple connection between common tools, Zapier. The reasoning is written into the discovery report.
Because the reasoning is written down, the choice can be revisited later. Because the workflow is documented, moving it to another tool does not mean starting over; the steps and exceptions are already on record.
The failed run is reported to its owner, its reason is written down and it is retried. If the retry also fails the workflow stops and a notification goes out. A workflow never stops quietly.
n8n runs on your own server and the data does not leave. If Make or Zapier is chosen, which fields go to that service is written during the build; sensitive fields never enter a cloud flow.
Yes. The workflows sit in your account or on your server and are handed over with their documentation. When you ask for a handover, training comes with it.
Steps such as payment, sending and deletion get an approval point. The workflow stops there and waits for a person; without approval it does not continue.
One, the one that takes the most time. Keeping the scope wide delays the first live run; a second workflow opens only once the first is measured and running.
Discovery takes a week. In that week we map the steps, the exceptions and the systems to connect, and choose the tool with its reasoning. The first live workflow opens within 4-8 weeks; the range depends on the number of steps, the systems to connect and the approval points. Your own timeline is written into the discovery report.
Two things are measured during discovery week: how long the step takes by hand and how often it repeats each week. Both are measured again once the workflow is running, next to the run success rate. The number of unnoticed outages is zero, because every failed run is reported.
The tools we use
Which workflow with which tool,and for what reason.
All six are in our toolkit. Each line below says when that tool comes out ahead; the reasoning is written the same way in the discovery report.
n8n
Runs on your own server, so data does not leave. Chosen for complex workflows and for anything carrying sensitive fields.
Make
Visual flows in the cloud; quick to build for mid-complexity, many-step jobs.
Zapier
A wide list of ready integrations; chosen for simple flows between common tools.
Our own code
Where no tool is enough, written on your own infrastructure with no ceiling on flexibility.
Airtable
A light data layer for small tables and approval lists; the team edits it directly.
Scheduled jobs
Overnight batches and reporting; large data is processed on your own server.
Related products
Where this is used
- Clinic automation: WhatsApp replies, reminders and one patient recordClinic automation is a set of rule-based workflows that run a clinic's repeated back-office work. It covers reply drafts for WhatsApp questions, appointment reminders, cancelled slots returning to the diary, post-visit messages and one patient record. ALTAI Digital builds these flows on the clinic's own rules. Drafts come from sources the clinic provides, and a person decides what is sent. The system takes no part in diagnosis or treatment. Health data is handled under GDPR Article 9. The first live flow opens within four to eight weeks.
- AI automation agency Netherlands: assistants and flows you can auditALTAI Digital is an AI automation agency serving small and mid-sized businesses in the Netherlands from Rotterdam and Istanbul. We build chatbots and voice assistants that answer only from your own sources, workflows documented with their triggers and error behaviour, and CRM setups that keep enquiries in one record. Where personal data is processed is written down at setup, in line with GDPR, and flows are handed over in your own accounts.
- AI automation agency Rotterdam: build it here, measure it hereALTAI Digital is an AI automation agency working from Rotterdam and Istanbul. One of our co-founders works from Rotterdam, so a discovery meeting can be held in person rather than only on a call. We build chatbots and voice assistants that answer from your own sources, documented workflows, and CRM setups that keep every enquiry in one record. Flows are handed over in your own accounts, with their documentation.
- AI automation agency UK: one working flow beats five demosALTAI Digital is an AI automation agency working with small and mid-sized businesses in the United Kingdom, from Rotterdam and Istanbul. We build chatbots and voice assistants that answer only from your own sources, workflows documented with their triggers and error behaviour, and CRM setups that keep enquiries in one record. Flows are handed over in your own accounts, and processing locations are written down under UK GDPR.
- AI automation agency: we hand over a working flow, not a demoAn AI automation agency builds, runs and measures the repetitive work inside a business with AI and automation. ALTAI Digital works in four areas: chatbots and voice assistants, workflow automation, CRM and enquiry handling, and websites with advertising. Assistants speak only from the business's own sources, workflows are documented, and everything is handed over in your own accounts. The first live flow opens in four to eight weeks.
- Law firm automation: intake, document drafts and deadline remindersLaw firm automation is a set of rule-based workflows that run a firm's repeated administrative work. It covers enquiry intake, document drafts from the firm's own templates, hearing and deadline reminders, client updates and a searchable document archive. ALTAI Digital builds these flows on the firm's own rules and templates. The system gives no legal advice; signature and decisions stay with the lawyer. Client data stays in the firm's own accounts and is not used to train models. The first live flow opens within four to eight weeks.
Terms explained
- AI agentAI agents take a goal, break it into steps and carry those steps out with tools. An agent does more than write text. It runs a search, reads a record, fills in a form or writes into another system. After each step it weighs the result and decides what comes next. Its permissions and its stopping rules are set during the build.
- AEOAEO stands for answer engine optimisation. It covers the work of writing a page so that answer boxes and AI assistants use it as a source. The method is to answer the question in the opening lines, write section headings as questions, and keep every claim checkable. Google states that this is not a separate discipline, and that the same SEO applies.
- n8nn8n is an open-source workflow automation tool that joins separate applications into one flow. A flow starts with a trigger and moves through steps called nodes. Each node does one job and passes the result to the next. There is a cloud version, but n8n also runs on a server the business controls. In that setup the data stays on that server.
Next step
The process first, the proposal after
In a short discovery call we map the process together and tell you plainly whether it is worth automating. If it is not, we say that too.
30 minutes · no preparation needed
Discovery call
Process maptogether
Automation candidatesprioritised
Measurement planin writing
Next stepclear
