Hybrid Search and RAG Retrieval with Elasticsearch
Vector search is not automatically the best retrieval strategy for RAG. In a small Elasticsearch experiment, we compare classic full-text search, semantic search, and hybrid search.
Latest on the blog
Hybrid Search and RAG Retrieval with Elasticsearch
Vector search is not automatically the best retrieval strategy for RAG. In a small Elasticsearch experiment, we compare classic full-text search, semantic search, and hybrid search.
Articles
Python, Web and AI Engineering
Technical articles on professional software development, architecture, Python and modern AI systems.
Understanding Tool Calling: How LLMs Use APIs and Other Systems
A Large Language Model can generate text, organize information, and suggest decisions. However, by itself it cannot load a customer record from a database, send an email, or create a calendar appointment.
Understanding AI Agents: What AI Agents Really Are
AI agents can use tools, make decisions, and perform tasks. But when is an agent more appropriate than a classic workflow?
Understanding Django Monitoring: Metrics, Logs, System Design
Many Django applications run for a long time "just fine." Requests come in, data is processed, responses go out. As long as nothing is obviously broken, the system feels stable.
Django Project Structure with a Focus on System Design
Django has long been a framework for quickly building functional applications. A few models, some views, a bit of template code, and a project was up and running. This way of thinking is now only partly valid.
Develop Django Faster with Mistral Vibe (2026 Guide)
Starting Django projects often follows the same pattern. Create project, make apps, define models, build views, wire URLs. The problem isn’t complexity but repeating the same steps. Especially for new projects or prototypes, this wastes time and focus.
Mistral Vibe Explained: What It Is and How to Use It Effectively
In recent years, numerous AI tools have emerged promising to make software development faster and easier. Many of these, however, remain superficial and mainly assist with small code snippets or simple tasks.
Services in Django: Clean Separation of Logic and Structure
Many Django projects start pragmatically. A few models, a few views, maybe a serializer – and everything works. The problem only becomes apparent later: The logic gets distributed uncontrollably throughout the code, making it increasingly difficult to understand where what actually happens.
How a Maintainable Django Project Is Really Structured
Maintainability does not arise by chance but through deliberate architectural decisions from the very beginning. This article demonstrates which layers, responsibilities, and project structures have proven effective in real Django projects—beyond tutorials and quickstarts.