AI-supported successor search for medium-sized companies

Development of an automatic system for investor identification

The aim of the project "AI-supported successor search for medium-sized companies" is to research and develop new machine learning models. Specifically, this means developing a model that can simplify and accelerate the process of identifying potential investors. This should make time-consuming manual steps partially obsolete, especially in connection with company succession in Mittelstand . The search for investors is one of the most time-consuming steps in the sale of a company, along with the preparation of the information memorandum.

Expected result

The planned software tool is based on the research and development of new machine learning algorithms, which are to be realised within the scope of this ProFIT application. Functionally, the tool is intended to support the analysts of Carl Finance in bringing together investors and companies for sale faster, more individually and more precisely, thus increasing the probability of a successful sale. Analogous to a Google search engine, the challenge is to automatically recognise relevance for the user from millions of data points. The user should be offered software that is as easy to use as possible, with the complexity taking place in the background. In particular, the process of creating the long list should be reduced from several days or weeks to a few hours or days.


With the support of the ProFIT funding programme, Carl Finance is researching two machine learning methods for the best possible implementation of the project:

The development of a neural text encoder based on Transformer architecture for domain-specific language processing in the context of corporate sales. The aim is to develop a new algorithm that can understand the intent in user search queries and automatically collect relevant data and create company profiles of potential buyers.



The development of a new meta-ensemble model that combines and fully integrates different machine learning models in a unified framework. The model should optimise itself and enable valid suggestions on potential buyers of companies based on individual input signals.


This project is co-financed by the European Regional Development Fund [ERDF].

Looking to invest?

We help you find the best companies in Mittelstand .

Schedule a call
close ×
Many thanks for making the appointment!
We will contact you by telephone at the agreed date.
Oops! Something went wrong while submitting the form.