As a seasoned software engineer with a passion for innovation, I bring a unique blend of technical interests and industry experience with me. With over two decades of hands-on experience in electronics and software engineering, I've had the privilege of working with top-tier companies like Varian (Siemens Healthineers) and Moser-Baer AG.
My journey began with an early fascination with technology, sparked by my first computer at the age of 10. Since then, I've witnessed the field evolve dramatically, with systems becoming increasingly complex. My experience has taught me that software development requires not only technical expertise but also a deep understanding of user needs, requirements and business processes.
Today, as a senior software engineer, I leverage my skills to drive innovation with a special focus on the use of AI.
When I'm not working, you can find me exploring the intersection of technology, philosophy and art, always looking for ways to push boundaries and create meaningful solutions - or I might just play a game 🎮.Â
There are many sophisticated Media Players for Linux with Jellyfin Server support. But some of them lack Chromecast and/or DLNA support - and if you are working from home with a coorperate device a native application is not an option. Therefore I'll try to integrate support to the docker version of Feishin ...Â
Feishin it is a robust, specialized client-side interface designed to exercise digital sovereignty over your personal media library. This open-source desktop application acts as a modern, functional frontend for established, self-hosted media servers (like Jellyfin or Navidrome). By decoupling the playback client from a centralized, proprietary cloud API (e.g., Spotify), Feishin allows users to maintain complete control over metadata, access patterns, and the underlying data.Â
In the rapidly evolving landscape of locally hosted Large Language Models (LLMs), performance often remains the primary bottleneck. While quantization and efficient kernel implementations have made running powerful models on consumer hardware a reality, a significant efficiency gap remains. This gap is bridged by speculative decoding, an advanced inference technique that allows local LLMs to operate with significantly greater throughput.
This post provides a technical summary of how speculative decoding works, the core mechanism, best practices for choosing a draft model, and the setup required for various tooling environments.
The openHAB community has released version 5.2.2. This release addresses multiple high-risk CVEs. If your current deployment is running on 5.2.1 or earlier, a review of your configuration is recommended to ensure the security of your setup.
The prevailing paradigm in LLM application development—relying on models to output descriptive text — is, from a systems engineering perspective, inherently unstable. When an LLM outputs a string — e.g., "billing" — this is not a guarantee; it is a hypothesis. Downstream systems must spend cycles to stabilize the output (regex parsing, string comparison, etc.). This uncertainty exposes the pipeline to linguistic variations.
A dynamic system using models from multiple vendors with different capabilities is vulnerable to such effects.
The Decision Model paradigm, formalized by typesafe.ai, addresses this constraint directly. These models do not generate text for aesthetic consumption; they are specialized, fast systems designed to eliminate ambiguity by returning deterministic, quantifiable outcomes: a choice, a confidence score, and a probability distribution.
Large Language Models, or LLMs, did not emerge from a single breakthrough. They are the result of several architectural innovations that changed how machines process language, learn from data and interact with external systems.
Modern AI applications are no longer just neural networks generating text. They can combine language models with retrieval systems, tools, memory, orchestration layers and agentic workflows.
This article explores that evolution, from token processing and recurrent networks to Transformers, retrieval and AI agents.
The project evolves from an MCP-centred experiment into a persistent, observable, and explicitly supervised runtime for multi-step software-engineering work.
Agent Orchestrator
Interaction using MCP App
The most important change in the project is not the addition of another model, tool, or user interface. It is a change in architectural responsibility. What began as an exploration of how agents and Model Context Protocol tools can cooperate evolves towards a supervised orchestration runtime. The current system no longer treats a model response as the workflow. Instead, it treats model calls as one part of a durable process that has an identity, an explicit plan, dependency-aware tasks, policy-controlled actions, recovery stages, operator handoffs, and observable results.
