Vertec and AI – Finding Opportunities in the Right Areas

At Vertec, we have been following developments in the field of AI with great curiosity since the launch of Chatgpt. We are enthusiastic about the pace of innovation, but as an ERP manufacturer, we are also keen on data security and applicable laws in our markets. In law firms, the adoption of AI tools has increased considerably. But we are also hearing increased demands and curiosity from other industries. Many of the developments indicate that all project- and mandate-oriented service providers will use AI in a beneficial way.

Where is Vertec current in relation to AI and what are our plans?

AI in and around Vertec

We already offer increasingly popular features in and around Vertec AI. If you use Vertec to enter expenses, you can save yourself a lot of work by using the AI-enabled receipt recognition. The service extracts information from images and PDF files and writes the information into the expense recording. Company with a lot of expenses benefit immensely from this first AI feature in Vertec.

In addition to the product, we offer an AI chatbot in the support area. The second generation provides customers with astonishingly good answers to a wide range of questions. The helpful answers are based on the daily updated Knowledge Base as well as further information on the Vertec model. The chatbot is active in hundreds of conversations per month, around the clock and in several languages. We monitor the response quality and continuously implement improvements.

Compiling Vertec by conversation: the AI agent in the configurator

Trying out Vertec has never been easier. We mean it literally. The new AI Agent in the configurator eliminates the tedious click-through of settings and options. Instead, simply describe what you need. The AI Agent guides you through our best practices in a structured way and sets up the right Vertec configuration. You get a Vertec that is precisely tailored to your industry.

The AI Agent knows the industry-specific requirements and helps you discover the right features with targeted questions.

6:52

<iframe allow="autoplay" allowfullscreen="true" class="embed-responsive-item" data-src="//www.youtube-nocookie.com/embed/J9Qs16Wu1aI?rel=0" frameborder="0" id="video-2382276" referrerpolicy="strict-origin-when-cross-origin" src="//www.youtube-nocookie.com/embed/J9Qs16Wu1aI?rel=0"></iframe>

Try ai agent in the configurator now

Future plans

At the Vertec User Conference 2025, we showed together with Dorian Selz from Squirro how an AI know-how management module can make knowledge and data from Vertec accessible via AI. Since then, the technology has evolved a lot. We have therefore reviewed and adapted our rate.

Our conclusion: For structured business data, the Model Context Protocol (MCP) is the leaner solution. Structured data can be addresses, projects, services or invoices, for example. Via MCP, an AI assistant directly accesses the current data in Vertec. The data does not have to be transferred to a separate system and processed there.

However, MCP does not replace RAG (Retrieval-Augmented Generation) in general. Sometimes large amounts of unstructured information need to be searched, such as meeting notes, emails or performance texts. For this, a RAG system remains the right choice. It helps to find information that is difficult to find with conventional search, and to make connections in large amounts of data visible. Read more in the blog post The vertec mcp server has outgone the rank of the rag system.

Handling structured data in the Vertec MCP Server

Structured ERP data such as quotes, invoices or business intelligence are, unlike text-heavy content, not so well suited for Llm. For this area, we are working on the possibility of making Vertec data accessible via MCP (Model Context Protocol). With a Vertec MCP server, we create the possibility for AI clients such as Claude from Anthropic to be connected to a Vertec instance. In the AI client, you then “talk” to Vertec. The MCP server provides certain features, which then gather the required data in Vertec. This enables queries like this:

• How many quotes are current open and what is the sum of the offered services?
• What is the amount of outstanding payments?
• How many services are not charged in project/case X?

The following demo shows the current state of development of the Vertec MCP Server.

12:13

Important: The MCP server should know instance-specific model information. Your adjustments in Vertec are not an obstacle in this AI use case, but a strength.