Where does our data run?

The workflows run on self-hosted n8n, in our German data centre or on your own infrastructure, with a proper data processing agreement. Where a language model is involved, only the content it actually needs goes to the model provider; we disclose exactly what that is, workflow by workflow. In our content pipelines that's product data anyway, not personal data.

Which LLMs do you use?

LLM-agnostic: OpenAI, Anthropic or Gemini. We choose by use case and budget, not by hype, and we switch models the moment a better one is available. The system and the workflows are unaffected either way; they belong to you.

Does this create a new lock-in?

No. n8n is open source, the workflows belong to you, and you stay portable. No platform pricing model that could change tomorrow.

What does an automation cost?

Less machinery, an honest price: you pay for building the workflow, not per-execution fees or corporate overhead. After the use-case analysis, you get a costing with disclosed assumptions, and an honest no if the automation doesn't pay off.

What can't automation do?

It doesn't replace strategy or judgement. We automate recurring, rule-based processes, not decisions that need a person. We define exactly where that line sits together with you.

What happens if a workflow breaks?

Retry logic, error handling and monitoring are part of the build, not bolted on afterwards. You get alerted before anyone else notices.

Does this pay off for an SME?

Especially there. No execution limits, no per-workflow fees, no need to add headcount for recurring tasks. An automation that still pays off six months in.

AI agents and workflow automation

AI agents on your infrastructure.

We build workflows and AI agents on n8n: open source, self-hosted, with a person in the decision loop. Data sovereignty instead of vendor lock-in, clear costs instead of opaque platform fees.

Discuss your project See our references

Our platform choice

n8n as the foundation, chosen deliberately.

n8n is our first choice for workflow automation and agentic systems. Open source, run on our own infrastructure, with full data control. Over 400 native integrations, custom nodes in JavaScript and TypeScript, and the ability to connect any system through its API.

What matters for SMEs: no execution limits, no hidden costs, no dependency on a platform that could change its pricing tomorrow. Something we believe in and work with every day.

Why n8n

Three reasons that make the difference.

Your data, your infrastructure

n8n runs self-hosted: in our German data centre, or directly on your own infrastructure. Your workflow data doesn't sit in someone else's automation cloud, and we disclose exactly what a language model gets to see, workflow by workflow.

Unlimited flexibility

Custom nodes in JavaScript or TypeScript, code you can drop straight into a workflow, any API you like. No vendor lock-in, you stay portable at all times.

Cost-efficient

No execution limits, no per-workflow billing. With self-hosting there are no recurring licence fees, so you keep full budget control even at high volumes.

How an automation gets built

1
Use case and data situation
- what runs manually
- which systems
- which data
2
Workflow design
- n8n flow
- error and retry logic
- sign-off stages
3
Connection
- APIs
- custom nodes
- LLM and RAG integration
4
Testing and sign-off
- dry run
- edge cases
- handover
5
Operations and monitoring
- logging
- alerting
- ongoing development

What we build with it

Automations that still pay off six months in.

Content pipelines with RAG: Product copy, blog scaling, SEO content from structured data sources. The same pipeline foundation as our SEO services.
CRM and email automation: Lead routing, segmentation, journey triggers. HubSpot, Mautic, Mailchimp, Shopware CRM or bespoke systems.
Customer support agents: Ticket routing, FAQ answers with RAG against your knowledge base, escalation logic. 24/7 coverage without adding headcount.
Multi-system synchronisation: ERP, shop, accounting, marketplaces, fulfilment. One pipeline keeps every system consistent, even when individual interfaces are flaky.
Webhook orchestration: Event-driven workflows across system boundaries. Retry logic, error handling, monitoring, all in the same tool.
Scheduled jobs and ETL: A replacement for cron hacks and ad-hoc scripts. Clean pipelines with logging, retries and clear ownership boundaries.

Agents, not just workflows

LLM-powered systems that work with your data.

The best-known example is our RAG content pipeline, which writes from real product data. Where an agent needs its own application logic, bespoke development comes into play.

LLM-agnostic: OpenAI, Anthropic or Gemini: we choose the model that fits the use case and the budget, and stay free to switch the moment a better one appears.
RAG integration: Agents pull from your knowledge base before every answer: product data, documents, CRM history. Less hallucination, more reliable answers.

From the field

An automation that actually runs.

Case Study

The self-driving content machine

An established mid-market online shop with several thousand specialist products that need real explanation, multiple country shops and a large trade magazine faced a scaling problem: high-quality, search-strong content, complete with imagery, internal linking and multiple languages, can't be produced by hand at the volume required. We built a fully automated content machine for it that generates, illustrates, translates and even voices articles, running autonomously round the clock.

Read the case study

All case studies

Clear boundaries

When we advise against AI.

Not every process deserves an agent. If a task only runs once a quarter, a checklist is enough. If a process has no clear rules, an LLM doesn't make it clearer, just wrong faster. And some problems get solved more cheaply by a good form than by any automation. We'll tell you that upfront, because the automation nobody needs is the most expensive one.

Where AI does hold up, our principle is this: a person orchestrates, the model is a replaceable tool. The workflows, the domain knowledge and the learning loop belong to you and get more valuable with every use, regardless of which model happens to be doing the work.

Frequently asked questions

What clients ask before an automation project.

Where does our data run?: The workflows run on self-hosted n8n, in our German data centre or on your own infrastructure, with a proper data processing agreement. Where a language model is involved, only the content it actually needs goes to the model provider; we disclose exactly what that is, workflow by workflow. In our content pipelines that's product data anyway, not personal data.
Which LLMs do you use?: LLM-agnostic: OpenAI, Anthropic or Gemini. We choose by use case and budget, not by hype, and we switch models the moment a better one is available. The system and the workflows are unaffected either way; they belong to you.
Does this create a new lock-in?: No. n8n is open source, the workflows belong to you, and you stay portable. No platform pricing model that could change tomorrow.
What does an automation cost?: Less machinery, an honest price: you pay for building the workflow, not per-execution fees or corporate overhead. After the use-case analysis, you get a costing with disclosed assumptions, and an honest no if the automation doesn't pay off.
What can't automation do?: It doesn't replace strategy or judgement. We automate recurring, rule-based processes, not decisions that need a person. We define exactly where that line sits together with you.
What happens if a workflow breaks?: Retry logic, error handling and monitoring are part of the build, not bolted on afterwards. You get alerted before anyone else notices.
Does this pay off for an SME?: Especially there. No execution limits, no per-workflow fees, no need to add headcount for recurring tasks. An automation that still pays off six months in.

Client voice

What our clients say about AI.

We've been working with total10 for many years, and it's always been fast and effective. t10 looks after and builds everything our various shops need. Lately, AI has been coming up more and more. What total10 makes possible there goes far beyond anything we'd imagined ourselves. Anyone on board now is years ahead of the competition. Anyone who isn't will be left behind.

Hendrik Pahl

Managing Director

Agent, workflow or ETL pipeline in mind?

Tell us what still runs manually in your setup today, or what's scattered across different tools. We'll show you how n8n brings it together cleanly. Reply within 24 hours.

Get in touch