AI implementation for the Mittelstand · Berlin / Brandenburg
VONTAI implements AI in owner-run German mid-sized companies: on site, in your real workflows, measured against one result we agree with you up front. Delivery, not advice.
The situation
of German companies with 50 to 249 employees did not use AI in 2025.
of those that considered AI but still do not use it name missing know-how as a reason. 59% name data-protection concerns.
Source: Federal Statistical Office of Germany (Destatis), ICT survey of enterprises 2025.
Remote teams and agencies know the process from a workshop, not from daily work. The solution fits the slide, not the desk.
When engineers or a single department write the brief alone, you get a wish list. What is missing is someone who makes product decisions.
Projects are measured by features shipped rather than by what changes in the business. Everything ends up “done” and nothing is better.
Our approach
AI-assisted development makes bespoke software affordable for a 100-person company. To make sure it lands in daily work, we follow four principles.
Our engineers work with you, not just for you. They see the real workflows, talk to the people who do the work and build together with them.
Every project has an accountable product lead who separates the essential from the nice-to-have, tests early with a prototype on your data and keeps the scope clear.
Before we build, we agree the one result that matters, with a baseline and a target. That is what we are measured by, and what you use to decide whether to continue.
A person approves every AI result. We use models processed in the EU and work with your existing ERP and tools instead of reinventing everything.
Services
Some companies need a solution for one specific bottleneck. Others want to build AI capability in-house. We do both, the same way.
For example email and order intake, reading documents, answering enquiries, matching data. From the first day on site to running operation.
For organisations that want their own AI capability: the structure, tools and routines your teams need to use and develop AI themselves.
We watch the real work, test a prototype on your data and agree the one result that matters. Credited against step 02.
An engineer works embedded in your team and builds against the agreed result. It ends with a clear decision: continue or stop.
All users, integration with your ERP and existing systems, on-site training. The solution becomes part of daily work, not another tool on the side.
We stay accountable: support, AI usage within agreed limits, continuous improvement. The terms are already written into the step-02 contract.
Durations are guide values. We set the scope for your project together in step 01.
Path B in detail
Each building block has a clear “done when …”. What you end up with is agreed up front, and use cases become the material your team learns on.
From our own practice
Three applications from our own companies. Each shows a building block we also use with clients.
Team, AI agents and finances at a glance, with the numbers that matter. This is what measuring impact looks like day to day.
I LIKE VISUALS GmbH · Tobias YorkAI sorts cases from shared mailboxes by issue and sentiment. Piloted on anonymised data from the company’s own operations.
flavity GmbH · Andreas SchwarzElim Camping, a booking platform for campsites, is being built with AI support at every step. It is how we build with your team, too.
Elim Camping · Eran NeubergerThe founders’ own applications, not client references. VONTAI is in formation.
Who we work with
We do our best work where decisions are made quickly and a solved bottleneck shows up directly in the numbers. Typical starting points:
Team
Product leader with more than twelve years in B2B software. At LivePerson he took a functions platform from zero to 80 enterprise customers, at Twilio he owned the Messaging and Conversations APIs, and as CPO of asellerate GmbH he built the product organisation and embedded AI in product and processes. At VONTAI he leads method and delivery.
Managing director of I LIKE VISUALS GmbH, Berlin. Twelve years of communication and strategy for purpose-driven organisations, including NABU. Runs his agency on a real-time dashboard for team, AI agents and finances.
Managing director of flavity GmbH. Uses AI in his own company’s daily work, for example to analyse service communication instead of sending customer surveys. Brings the view of an owner who has to make the numbers work himself.
For delivery we work with experienced engineers and specialists from our network, brought in to fit each project.
Contact
Describe in two or three sentences the workflow that costs you the most time. We will reply personally and tell you honestly whether and how AI can help there.