Intelligently Connected

Use cases that were considered utopian or even impossible just five years ago have become a reality with the widespread availability of AI systems. Traditional business software systems such as ERP or CRM feature robust, deterministic algorithms and structured, rich domain data. When combined with the advanced capabilities for data analysis and aggregation that AI offers, users can benefit immensely from this synthesis—provided that the right use cases are addressed, ones that deliver real added value to the end user. Using AI just for the sake of AI is hype; using AI to genuinely increase efficiency and benefit is smart.Translated with DeepL.com (free version)

Multidata chooses the latter path.

AI-assisted order creation

Orders that arrive by e-mail are a constant nuisance for customer service because every e-mail costs time, ties up resources and is a potential source of error. Automatic processing of such e-mails has so far failed due to the complete unpredictability of what information is sent, where it can be found and in what format.

MD-Premium AI couples two already proven functions (e-mail sync via Microsoft Graph, order import from external webshops) with an AI to achieve this automatic processing. The AI analyzes the email to see if it is an order, extracts all relevant information, and creates a standardized data format that can be imported from MD-Premium.

The system allows final control and approval by a human decision-maker and offers convenient tools to resolve problems that have arisen during import.

Used consistently, this feature can save your customer service hours of work every day, which are freed up for higher-value activities.

MCP Interface

With the help of Agentic Workflows, you have virtually unlimited possibilities to combine AI models and classic IT systems to perform complex tasks. In this context, IT systems such as ERP, HRM or document archiving usually take on the role of a data source or provide specific functions that an AI model itself cannot master.

The MCP interface from MD-Premium guarantees you universal connectivity of your software to the AI world. Thanks to the MCP standard defined by Anthropic, you can make accessible functions and data that can be called by AIs in a targeted manner.

Thanks to the flexible adaptability of the MD-Premium AFW, you can precisely determine which function calls are made available to an AI and which parameters are required in just a few simple steps.

Whether you have AI agents deeply integrated into your process flows or only sporadically use AI models to support your work... with the MCP interface of MD-Premium, your business data is available "promptly".

Semantic vector search

AI models reach their limits when they have to work with a larger database on which they have not been trained. Since training your own model is hardly justifiable in terms of effort, the semantic vector search offers an economical alternative.

MD-Premium uses special data structures (semantic vectors) to give natural-language texts a semantic rating. On the basis of this semantic evaluation, queries can be better understood and the database can be searched much more precisely for possible hits. This created a pre-selection, which in turn can then be handed over to a generic AI for the final formulation of the answer.

The big advantage of vector search is the resource-saving process of vectorizing data, which requires only a fraction of the time and energy needed to train AI models. Thus, this approach is excellently suited to feed non-public and constantly changing data to an AI for processing.

Semantic vector searches can be used in many places in MD-Premium, e.g. in the documentation module, in the ticket system or in the PIM database.