Posts

Software documentation today is more than just a mandatory task. It is part of the development process and must be as agile, versionable and automatable as source code itself. Classic office documents or proprietary wikis quickly reach their limits. Changes are hard to trace, collaboration is cumbersome and integration into CI/CD pipelines is hardly possible. AsciiDoc and Kroki address exactly these points. Both tools embody the Docs-as-Code approach, where documentation is treated like source code. AsciiDoc provides an expressive text-based language for technical documentation that can be managed in Git and processed automatically. Kroki complements this with the ability to embed diagrams directly in the document and render them automatically. This allows developers to use diagrams without needing to install additional tools. ...

Introduction: Why control over data determines AI success Artificial intelligence does not emerge in a vacuum, it relies on data as fuel. A neural network without a broad and high-quality data foundation can neither process language nor identify objects or derive meaningful recommendations. For companies, this results in a clear consequence: whoever loses control over their own data leaves the crude oil of the digital economy to external platforms. The concept of data sovereignty describes exactly this factual and legal dominion over one’s own data holdings and goes far beyond classical data protection. While data protection primarily aims to safeguard individuals’ fundamental rights by protecting their information, data sovereignty represents a strategic and economic question. It is about who owns data and who is allowed to use it in which way. This determines whether data becomes a valuable competitive advantage or flows unnoticed into the value chains of others. ...

In machine learning there are many paradigms and methods from supervised and unsupervised learning through semi and self supervised to transfer learning and reinforcement learning. To keep an overview I have created a mindmap as a memory aid. On the first level are the learning types such as supervised learning, then the task areas, followed by the thematic subgroups and finally concrete methods. The visualization makes clear how broad the field is and facilitates the classification of the individual methods. It shows only a subset of the possibilities. Many other methods exist, I have focused here on the most frequently used and established in practice methods. ...

In Part 1 we saw how crucial clean document preparation and thoughtful chunking are to the quality of Retrieval Augmented Generation. These basics form the starting point for a whole range of further optimizations that shape the entire process. In Part 2 we continue the series and focus on the next building blocks that build on this foundation and further develop the use of RAG in the enterprise. Embedding Domain-specific Embeddings Domain-specific embeddings mean that vector representations of texts are not generated with generally trained embedding models, but with models adapted to the technical language and content of a specific industry or company. General models are trained on very large, unspecific text corpora, including books, websites, Wikipedia and other sources. They understand everyday language and many standard concepts, but often miss the nuances in, for example, legal contracts, technical manuals or medical reports. Domain-specific embeddings are created either by fine-tuning an existing model with data from the respective domain or by training a custom model on a corpus of internal documents, guidelines, protocols and manuals. ...

Retrieval Augmented Generation, or RAG for short, combines the power of language models with a company’s specific knowledge. The approach makes it possible to incorporate internal documents and data into responses in a targeted way without losing control over one’s own information. As a result, RAG is increasingly seen as a key technology for deploying language models securely and with data sovereignty. In practice, however, it quickly becomes apparent that simple vector search in combination with an LLM is not sufficient to achieve truly consistent and high-quality results. To fully exploit the potential of RAG, additional methods and optimizations are necessary. ...

Introduction Large language models (LLMs) like Llama 3.x are trained in an elaborate pretraining process on massive amounts of text. This process typically takes place on specialized hardware such as GPUs and TPUs, which are optimized for parallel computation of large neural networks. After pretraining is complete, the model parameters are frozen and can no longer be directly changed during normal operation. This means that you cannot simply “correct” the model or reprogram it with simple interventions. Content such as facts about historical figures is not stored in individual, explicitly addressable neurons. Instead, such information is statistically distributed across the entirety of the model weights. This makes targeted modifications considerably more difficult, as there are no clearly identifiable storage locations for specific facts. ...

Introduction With the advent of large language models (LLMs) such as GPT, many people wonder how to provide these models with structured, precise information. Although LLMs are capable of answering questions very convincingly, many of their answers are based solely on statistical language probabilities, not on logical inference or explicit factual knowledge. This is where the use of an ontology offers systematic added value. In the following article, a fictional mission in the „Lord of the Rings“-Universe is used to demonstrate how an ontology can support an LLM in answering complex questions. ...

Introduction Unstructured.io is an open-source framework for the structured preparation of unstructured documents such as PDFs, Word files, HTML pages, or emails. Its goal is to extract semantically usable content from these heterogeneous formats—such as headings, paragraphs, tables, or lists—and convert it into a unified, machine-readable format. The main use case lies in preparing text data for downstream AI processing, particularly for systems with retrieval-augmented generation (RAG). The typical application is document analysis, knowledge management, or preparing inputs for embedding models. Multiple processing steps are employed. These four steps form the core of the Unstructured.io pipeline and are executed in every regular use of the library. ...

Introduction Retrieval Augmented Generation (RAG) is a technique in natural language processing (NLP) where a language model is combined with external knowledge to produce better and more precise answers. A language model like GPT is queried not only on its internal knowledge (training) but also receives context-specific information from an external knowledge source, e.g. a document collection or database. This article explains the structure and development of a RAG pipeline as part of a learning project. The goal was to develop a system that processes the content of a PDF document and enables an interactive chat to ask questions about this document. The application was born from the desire to practically understand the functionality and interplay of the individual components of a RAG application. ...

Introduction Stanza is an open source NLP library from Stanford University based on modern neural networks. It enables comprehensive linguistic analysis of texts in over 70 languages. Stanza’s goal is to provide a complete pipeline system that includes all common processing steps: tokenization, part-of-speech tagging (POS), lemmatization, syntactic analysis (dependencies and constituency) as well as Named Entity Recognition (NER). Stanza is suitable both for research purposes and for production applications, such as text classification, information extraction or preprocessing texts for retrieval-augmented generation (RAG). The models are pretrained but can also be fine-tuned. Internally, Stanza is based on the PyTorch framework. ...