Skip to content

Understand first, then build.

From the first conversation to a solution in everyday use: four stages from a single source. One contact instead of three providers who have to coordinate with each other.

Every stage ends with something tangible in your hands.

  1. 01

    Orientation

    A conversation with one of the founders, by video or at your premises. We tell you frankly where AI pays off and where existing tools are enough.

    What you have afterwardsA list of possible starting points and a recommendation for the next step.

    Orientation meeting

  2. 02

    Analysis

    We look at the process where it happens: conversations with your people, work on site, real cases.

    What you have afterwardsMeasured baseline figures, a business case and a roadmap. Plus a clear statement of what does not pay off.

    Process analysis · one process · Area analysis · several business areas

  3. 03

    Implementation

    We build the solution, connect your systems and test it on your real cases. First it runs alongside, then we train your team.

    What you have afterwardsA tested, introduced and signed-off solution with documentation.

    Assistant · Workflow · AI agent within a fixed framework

  4. 04

    Operation

    We look after the solution in day-to-day operation: we monitor it, fix faults and adapt it to new models and interfaces.

    What you have afterwardsA monthly report on quality, usage and costs, and a dedicated contact.

    Operations · Ongoing development

Where do you stand at the moment?

Some know exactly which process is holding them back. Others have plenty of ideas but no order of priority yet. Choose the statement that fits best.

‘We do not yet know where to start.’

Your starting point

01 Orientation

Offer
Orientation meeting
First
A conversation about your processes, your ideas and what is already in place.
Afterwards you have
A list of possible starting points with a recommendation for the next step.

‘Our team uses AI, but everyone does so on their own and without any noticeable benefit.’

Your starting point

01 Orientation

Offer
Orientation meeting on AI use in the team
First
Together we look at what your team uses AI for today and with which tools. Openly, and without assessing individuals.
Afterwards you have
A proposal for turning individual use into shared work: clear rules, one shared access point and the tasks that lend themselves to a fixed process.

‘One particular process is costing us too much time.’

Your starting point

02 Analysis

Offer
Process analysis
First
Conversations with the people who carry the process every day, and an evaluation of real cases.
Afterwards you have
Baseline figures, assessed options, benefits and costs, and a roadmap.

‘We already have a solution, but it is not delivering.’

Your starting point

02 Analysis

Offer
Process analysis focusing on the existing solution
First
We check whether the problem lies in the technology, the process or the roll-out.
Afterwards you have
A clear recommendation, with reasons: improve it, replace it or leave it.

‘The project is clear and we want to implement it.’

Your starting point

03 Implementation

Offer
Implementation after a review of your preparatory work
First
We review your preparatory work and define the target process and acceptance criteria.
Afterwards you have
A tested, introduced solution that your team works with.

AI has arrived in mid-sized businesses. Its value has yet to follow.

Most companies now use AI. For the most part it helps individuals with writing and searching, and rarely entire processes.

57%1
of companies with 20 or more employees use AI.In 2024 the figure was 20%, in 2025 36%.
59%2
of AI users, by their own assessment, are not exploiting its potential at all.Only 3% make fairly extensive use of it.
27%3
of AI users use it to automate processes or decisions.35% use it to write texts.
Sources
  1. 1 Bitkom, press release of 14/09/2026, companies with 20 or more employees (n = 603), comparison figures from the same series. bitkom.org · last checked on 10/11/2026
  2. 2 Bitkom, press release of 14/09/2026, companies using AI with 20 or more employees. bitkom.org · last checked on 10/11/2026
  3. 3 Destatis, 2025, companies using AI with 10 or more employees. destatis.de · last checked on 10/11/2026

What this looks like in practice. Shown on processes everyone knows.

Six examples from everyday work in mid-sized companies. Each one shows what the AI takes on and where your team decides.

Sorting documents

Today

  1. Your inbox
  2. Open, rename, file, re-key.
  3. ERP

Someone opens every PDF, renames it, files it in the right folder and types the details into the ERP.

Afterwards

  1. Your inbox
  2. FilingAI alone
  3. Details for the ERPAI drafts, a person checks
  4. Awaiting checkAccounts
  5. ERP

The filing sorts itself. Your team checks what gets posted and what was uncertain.

Capabilities: Classifying, Extracting · Demonstration with sample data · Lindtaler Anlagenbau GmbH is fictitious

Use cases: Sorting documents

Answering email enquiries

Today

  1. Enquiry
  2. Search first, then reply.P30 manualEarlier casesP30 spare parts list

Anyone answering an enquiry searches through manuals, earlier cases and spare parts lists.

Afterwards

  1. Enquiry
  2. CheckIs the approved knowledge enough?
  3. DraftAI drafts, a person checks
  4. Awaiting approval by your service team

If the approved knowledge is enough, a draft with sources is ready. If not, it says what is missing. Your service team checks and decides what goes out.

Capabilities: Classifying, Searching, Generating · Example with fictitious data

Use cases: Answering email enquiries

Development status

  • Demonstration shows what is possible, using sample data
  • Pilot with your cases, first in testing, then in shadow operation
  • Operation runs reliably day to day
All use cases

What you can rely on.

  • A fixed price before we start.Scope, result, price and your contribution are set out in the proposal. Changes are agreed in writing before they create additional work.
  • You decide after every stage.Each stage is commissioned separately and ends with a result. Whether and how to continue is then up to you.
  • The results are yours.Even if you do not implement with us. Any implementation partner can start from the analysis without collecting the data again.
  • Your team stays in control.Consequential decisions stay with your people. A solution only gains more autonomy step by step, measured and with your approval.
  • Your data stays protected.In operation, the language models run in the EU. Leading-edge models only with your approval. No provider trains on your data; fine-tuning for your task only takes place with your approval.
  • No lock-in.If our collaboration ends, we hand over data, documentation and the source code developed for you. The solution keeps running without us.

What you will want to know before the first conversation.

What does it cost?

That depends on the project, which is why we do not quote flat rates here. After the first conversation, you receive a proposal with a fixed scope and a fixed price.

How much of our team’s time will it take?

We need one dedicated contact person and conversations with the people who carry the process. We tell you how much time is needed before we start, so that you can plan for it.

Do you need access to our systems, and will it work with our ERP?

For the analysis, we do not need any access. We work with conversations, observation on site and sample cases. For implementation, we build on what you already have instead of introducing a new system, and connect whatever the solution needs, in agreement with your IT department.

What makes you different from an IT service provider or an agency?

An IT service provider looks after your IT and your software; an agency usually builds whatever it is commissioned to build. We start with the process: we check whether AI pays off, build only what makes financial sense and then look after it in day-to-day operation. We coordinate with your IT department and your IT service provider, and we will also recommend their solution if it is a better fit.

What if AI does not pay off for us?

Then we say so. Sometimes existing software, a simple rule or a change in the process is enough. That is a result too.

Is your question not listed?

Ask us directly. We reply within two working days.

Ask a question

Which task takes up most of your team’s time?

Bring a specific task or an initial idea. You do not need a finished project brief.

Your first conversation is directly with one of the founders: Julius, Luis, Max or Quentin.

Prefer to write directly? kontakt@omnerio.com Two or three sentences about your task are enough.