05
AI automation, productized
AI automation is a packaged engagement: a chatbot that answers your customers, or a workflow that moves data and triggers actions without a person in the loop. The scope is small enough to price up front, and you can estimate it yourself before we talk. Deeper agentic or retrieval-heavy work routes to a separate quote.
Pricing
€1k-€10k
Indicative range for a productized chatbot or workflow. Deeper agentic work is quoted.
What's included
- A chatbot for your site or inbox, rule-based for fixed questions or NLP-backed when answers vary.
- A workflow that connects the tools you already use, with webhook triggers so steps fire on their own.
- A scoped plan with a fixed price and a fixed end date, agreed before we build anything.
- Logging and a clear fallback, so when the bot is unsure it hands off to a human instead of guessing.
- Handover-ready setup you own and can change, with no monthly fee back to us.
How it works
- 01
Pick the lane
A short call to decide which it is: a chatbot, a workflow, or both. We also name what is out of scope, so the price stays fixed.
- 02
Fixed proposal
We send back the exact build, the price, and the end date. If the job is deeper than this tier, we say so and quote it separately instead of stretching a fixed scope.
- 03
Build and connect
We wire the bot or the workflow to your tools, test it on real cases, and set the fallback for when it is unsure.
- 04
Handover
You get the working setup, the access, and a short walkthrough. You can edit it without us.
Most of these run in days to a few weeks. Fixed scope, fixed price. Deeper agentic work is quoted on its own.
Common questions
How is this different from your AI integration service?
This is the packaged tier: a chatbot or a workflow with a scope small enough to price before we start. AI integration is the deeper engagement, grounded in your data with evaluation and retrieval. If your job needs that, we point you there.
Rule-based or NLP chatbot, which do I need?
Rule-based fits a fixed set of questions and is cheaper to run and easier to predict. NLP fits when customers ask in their own words. We pick on the call, based on your actual questions, not on which sounds more advanced.
Will the bot make things up?
We scope it to answer from what it knows and hand off to a human when it does not. For answers that have to come from your own documents and data, that is retrieval work, which sits in the AI integration service, not this tier.
What does the workflow part actually do?
It connects tools you already use and runs steps on a trigger, like a new form entry or a webhook, so a person stops copying data between apps. We keep it to the simple and medium cases here; anything that needs custom logic gets quoted separately.