ViraFlow TecDoc

netwyk ViraFlow enables the automated processing and analysis of incoming emails, including file attachments, through configurable AI workflows. Users can use a browser-based interface to define individual prompts and process logics that determine how content is interpreted, analyzed, and further processed.

Incoming emails are received by the system, relevant content and attachments are extracted and then processed by the defined AI workflows. Based on the stored instructions, the application generates analyses or formulated answer suggestions. This allows complex, document-based processes to be flexibly automated and adapted to specific use cases.

The system is able to process PDF, Excel and Word documents.
If you are working with Excel files, the system will only consider the first worksheet.

For successful processing, transmitted documents must be in a freely accessible and machine-readable form. Password-protected, encrypted or otherwise secured files against automated readout cannot be processed by the system and are therefore excluded or lead to an aborting of the workflow.

In addition, the processing of purely image-based content is not guaranteed. This applies in particular to scanned or photographed documents as well as embedded graphics without a searchable text layer. Such content cannot be reliably analyzed and can lead to incomplete results or workflow abandonment.

Data (as of 2026-04-16)

• License „S“:
  • Pages: max.
  • Excel cells max.
• License „M“
  • Pages: max.
  • Excel cells max.
• License „L“
  • Pages: max.
  • Excel cells max.
• License „M“
  • Pages: max.
  • Excel cells max.
• License „PPU“(Pay per Use)
  • Pages: max.
  • Excel cells max.

The platform LLMs for processing, analyzing, and generating content. These models enable the semantic interpretation of unstructured data such as emails and documents, as well as the creation of contextual responses based on user-defined workflows and prompts.

A distinction is made between pretrained and trained or fine-tuned LLMs.

Pre-trained models are used as standard. They have been trained on large, general-language and cross-domain datasets and thus have a broad basic understanding of linguistics and semantic. They form the basis for general language processing tasks such as text analysis, summarization, and response generation.

In addition, additionally trained or fine-tuned models can be used. These are specifically adapted to specific use cases or corporate domains to increase the accuracy and relevance of the results in the respective context.

I am text block. Click edit button to change this text. Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.