Christian Krudewig

Coding agent in a box

Sun, 18 Oct 2026 00:00:00 +0000

Coding agents have changed how we build software, but they’ve also created new risks. This post describes a response to the one that worries me most: an agent compromising the systems I have access to, or leaking sensitive information.

Shaping Engineers in the Age of AI

Wed, 16 Sep 2026 00:00:00 +0000

This is the story of how AI coding agents changed the informal way we teach engineering on the job, and what we can do to keep turning juniors into seniors.

Designing HTTP APIs for slow operations

Thu, 11 Jun 2026 00:00:00 +0000

We got used to having everything on the Web, and so many backend operations became HTTP APIs. That’s also the case in the areas of AI and optimization, because it gives an easy-to-understand interface for clients that can be used without having to learn new technologies. There’s a problem, however: Some of these operations are very slow, and at times systems become a brittle mess of regular timeout hits, latency optimizations, and support issues.

Services

Mon, 01 Jun 2026 00:00:00 +0000

Ingress Nginx retires: it's time for Kubernetes Gateway API

Mon, 09 Feb 2026 00:00:00 +0000

The Ingress Nginx controller is retiring next month, in March 2026. What comes next? I think it should be the Kubernetes Gateway API - and in this post I will explain why.

Self-Contained Teams: How to Keep a DevOps Culture in a Growing Organization

Thu, 22 Jan 2026 00:00:00 +0000

When projects take longer despite your organization gaining more expertise, when developers share frustration about “process hindering progress”, the problem may not be the people or the technology. It may be time to look at team boundaries.

Analyzing One Year of Solar & Heat Pump Data

Tue, 06 Jan 2026 00:00:00 +0000

It is now 1.5 years since we replaced our gas boiler with a heat pump and installed a solar panel system. We had heard a lot of horror stories about misconfigured heat pumps and the excessive costs that they may cause, so I invested into detailed planning and a custom simulation of the photovoltaic system.

Storing evaluation data for AI projects

Wed, 17 Dec 2025 00:00:00 +0000

When building ML systems, test and evaluation datasets are as critical as the code itself. But where should you store them? The answer depends on size, sensitivity, and how often they change.

Simulating Home Solar with Python

Sat, 01 Jun 2024 00:00:00 +0000

Ever wondered if a solar installation would pay off? I’ll show you how to find out with Python.

When we had ordered a new heat pump as replacement for the old gas boiler, we assumed that a solar installation would also be a good idea. Not only because we could make a little contribution to the rise of renewables, but with the additional electricity demand it was likely to pay off financially, too.

Python packaging unboxed

Fri, 23 Feb 2024 00:00:00 +0000

Python packaging has a reputation for being confusing. With tools like setuptools, pip, poetry, flit, pdm, and twine, it’s easy to get lost. This post demystifies the packaging workflow from the import system to Docker deployments.

About

Mon, 01 Jan 2024 00:00:00 +0000

About me

From economics through operations research to software and AI engineering, I’ve always focused on solving real-world problems with technology. I’ve built and scaled machine learning systems, designed robust software, and transformed high-maintenance systems into simpler, more dependable solutions. My passion lies in exploring new tools and methods, but I always prioritize functionality, reliability, and tangible impact:

Python 3.11 thirty percent faster indeed

Thu, 10 Nov 2022 00:00:00 +0000

Two weeks back the new version 3.11 of Python was released with the promise of a significant speedup. Let’s test it!

For fun, I’m maintaining a small benchmarking project. There, a couple of computational tasks are solved with Python and with different speedup methods such as Cython, Numba and C extensions written in C++ or Rust. Main purpose for me was to figure out which method has the greatest effect and how good the user experience is.

Locating memory leaks in a Python service – Part 2: C-Extensions

Mon, 12 Sep 2022 00:00:00 +0000

Building up on last week’s issue about diagnosing memory leaks in Python, this blog post will show how to identify memory issues in native extension modules.

From [the first posting]({% post_url 2022-09-04-locating-memory-leaks-in-services-part1 %}) we have the tools to clearly diagnose a memory leak in pure Python. So let’s proceed to the next level and add some native code to the picture. It is one of the strengths of Python how easy it can be extended with compiled code e.g. in C, by so called “extension modules”. This is a key factor which made Python the first choice for data science: The heavy number crunching can be done by optimized routines and the program logic is written in the high-level Python language.

Locating memory leaks in a Python service – Part 1: Pure Python

Sun, 04 Sep 2022 00:00:00 +0000

Even in a garbage-collected language like Python it is possible to create applications with memory leaks. This means that the program keeps memory reserved vor variables which are no longer needed. So the memory could be freed, but it isn’t. Especially in long-running applications this can lead to a huge memory consumption over time, up to a point where the hardware limit is reached. In a series of two blog posts I’ll show how to debug such memory issues in a Python web service.

Testing a Flink Stateful Function Service

Fri, 15 Apr 2022 00:00:00 +0000

With Flink Statefun a powerful, polyglot data streaming framework has entered the real-time data processing landscape. It allows to break down a streaming pipeline into individual microservices which can each take care of one or more tasks. This post is about how to test these services individually without executing the whole pipeline.

Machine Learning Models Light as a Feather with ONNX

Thu, 09 Sep 2021 00:00:00 +0000

Have you ever faced the challenge to create a fast and lightweight web service out of your machine learning model? In this case you are maybe annoyed by the overhead introduced by all the machine learning dependencies from the training environment that you also need to carry over to the production environment. This post tests an approach which claims to make it possible to leave most of this behind and share the blank model in a way that it can even be used across multiple programming languages.

Envoy sidecar configuration generator

Thu, 06 May 2021 00:00:00 +0000

This post contains a configuration file generator for an envoy reverse proxy with all the bells and whistles. If you are looking for a working base configuration with the most important features of Envoy as sidecar proxy, this is the place to get one.

The Throughput Paradox – Why a Machine Learning API Needs Load Shedding

Sun, 02 May 2021 00:00:00 +0000

Especially compute-intensive services like a machine learning inference API can heavily suffer from queueing effects. I will demonstrate this with response time measurements of small demo service and show how to deal with this using load-shedding

Getting Started with an Envoy Sidecar Proxy in 5 Minutes

Sun, 18 Apr 2021 00:00:00 +0000

This post shows the basic setup of Envoy as a reverse proxy in a sidecar container. It will show a typical setup of a small web API accompanied by an Envoy as sidecar which does nothing else than just forwarding requests to the backend. That’s the right starting point for adding more features and experimenting with the setup.

Multicluster Application Monitoring with Prometheus

Mon, 22 Feb 2021 00:00:00 +0000

Prometheus is a popular choice for application monitoring. It is easy to set up and can be deployed alongside the applications inside a Kubernetes cluster. However, when you cannot run your main Prometheus instance within the same cluster it becomes a bit more tricky. At work we recently set up Prometheus monitoring for a web service which is deployed to independent kubernetes clusters in different regions. The official documentation is a bit sparce on this topic. In this post I will show what I learned about combining the metrics of applications running in separate Kubernetes clusters.

Hosting a Blog the GitOps Way

Tue, 01 Dec 2020 00:00:00 +0000

Thanks to static site generators, a blog can easily be driven and controlled by markdown files in a git repository. This way, it is not necessary to set up a content management system including a database ever worrying to end up hosting yet another hacked Wordpress instance. Instead one can focus on the content and publish by simply pushing to the main branch.

Contact

Mon, 01 Jan 0001 00:00:00 +0000

Let’s talk about your project, or just exchange ideas. Pick the option that works best for you.

Book a meeting

Legal Notice

Mon, 01 Jan 0001 00:00:00 +0000

The German version is legally binding. — Impressum (Deutsch)

Information pursuant to § 5 DDG

Christian Krudewig
Marienstraße 49
53773 Hennef
Germany

Contact

Email: contact@krudewig.dev
Book a meeting: Online appointment via Calendly
VAT ID: DE463655358

Privacy Policy

Mon, 01 Jan 0001 00:00:00 +0000

This is only a translation. The German version is legally binding: Datenschutzerklärung (Deutsch)

1. Controller

The controller responsible for data processing on this website is:

Christian Krudewig
Marienstraße 49
53773 Hennef
Germany