Coding Agents V: Why Bubblewrap Wraps Agents
Bubblewrap does in 30 lines what Docker does in 30 layers. A deep dive into the Linux sandboxing technology that wraps Coding Agents in battle-proofed security.
Coding Agents IV: Enterprise — Ban, Buy, or Build?
For companies, Coding Agents are both an opportunity and a risk. Part 4 of our series maps out the three strategic options: ban, buy, or build — and why understanding the technology is non-negotiable for leaders.
Coding Agents III: Sandboxing & Best Practices
Pi has no built-in boundaries — it operates with full user permissions. Part 3 of our series shows how to run Coding Agents safely using Dev Containers, and shares hard-won best practices for productive agent workflows.
Coding Agents II: Prompts, Tools & Skills
A Coding Agent is only as good as its instructions and its tools. Part 2 of our series demystifies System Prompts, Tools, Skills, and why Pi deliberately keeps them minimal.
Coding Agents I: Beyond Chat
Coding Agents don't just suggest code — they write, execute, and iterate on it autonomously. Part 1 of our series on practical Coding Agent adoption.
Escaping CUDA Dependency Hell: Why Virtual Machines are the Ultimate Solution for Deep Learning
Managing PyTorch and CUDA versions across multiple projects often leads to 'dependency hell.' We explore why standard isolation tools fail and how VM-based solutions provide the ultimate fix for [**autoVI**](https://autovi.de/) and other deep learning projects.
DeepTech Series Part 4: 3D Leadership
Why 3D leadership is essential for DeepTech projects to succeed. The final part of our series on Germany's DeepTech innovation engine.
Accelerating Time Series Exploration: From Naive Pandas to Polars GPU
At Palaimon, developing autoStreams involves processing massive amounts of time series data, from identifying rare events in sensor data to aligning high-frequency financial streams.
Announcing DIN SPEC 91531: A Digital Standard for Recycled Plastic Packaging
Announcing DIN SPEC 91531, a new digital standard for recycled plastic packaging that provides a common language for the industry.
Choosing the Right Time-Series Architecture: ArcticDB vs. ClickHouse and TimescaleDB
Why serverless DataFrame databases like ArcticDB are the superior choice for initializing real-time streaming in [**autoStreams**](https://autostreams.de/) compared to traditional OLAP engines.
Scaling Similarity: The Power of Approximate Nearest Neighbors (ANN)
At Palaimon, our dive into Approximate Nearest Neighbors (ANN) was born out of a real-world engineering challenge while improving autoVI, our automated visual inspection solution.
DeepTech Series Part 3: The Alignment Gap
When technical excellence doesn't find its way to value. Part 3 of our series on Germany's DeepTech innovation engine.
DeepTech Series Part 2: The Prototype Wall
When laptop success crashes into hardware reality. Part 2 of our series on Germany's DeepTech innovation engine.
New Publication: AI in Pre-Hospital Tele-Emergency Medicine
New publication on AI in pre-hospital tele-emergency medicine, presenting research and applications in emergency response.
DeepTech Series Part 1: The Lab Trap
Why brilliant research often fails to survive the 'Real-World Atmosphere'. Part 1 of our series on Germany's DeepTech innovation engine.
Bridging Theory and Practice at APPIS 2025: Exploring Sequential Optimal Design
I recently had the pleasure of attending the 7th Applications of Intelligent Systems (APPIS 2025) conference, held at the University of Las Palmas de Gran Canaria (ULPGC).
Enhancing Python Project Productivity: The Advantages of Mamba for Environment Management
The multitude of Python environment management choices can leave even senior Python developers unsure of the best tool usage. This article describes how Palaimon reduced costs while increase productivity by choosing to use mamba/micromamba for environment management.
PyCon US 2023
PyCon US 2023 recap: tutorials on Jupyter notebooks, PyPI organization accounts, and keynotes by Carol Willing and Guido Van Rossum in Salt Lake City.
CityCount Feasibility Study Completed
With project CityCount we have demonstrated that an AI-based traffic counting of non-motorized traffic participants is feasible in a privacy-aware manner.
K fair -- all about plastics
In October 2022 we visited again the K fair in Düsseldorf to learn more about the newest trends in the plastics industry.
Kick-off KIOptiPack: AI Application Hub Plastic Packaging
Sustainable Circular Economy through Artificial Intelligence - KIOptiPack project kickoff.
Ipyannotator in The Journal of Open Source Software
Ipyannotator was published in The Journal of Open Source Software in August 2022, an important step in promoting the open source and highly customizable annotation framework for data-centric AI.
Kick-off KIT² -- better medical emergency response with AI
Kick-off meeting for KIT² project to develop better medical emergency response with AI technologies.
Python Nordeste 2022 Recap
From August 26th to August 28th 2022 I attended the Python Nordeste, the biggest Python event in northeast of Brazil.
EuroSciPy 2022
Palaimon got invited to talk about mathematical optimization involving uncertainty at the EuroSciPy 2022. Beside valuable exchanges with other experts, fascinating new python packages and their application, we gathered other interesting insights and values.
Literate programming with Jupyter notebooks + nbdev + Quarto
Exploring literate programming with Jupyter notebooks, nbdev, and Quarto for reproducible scientific workflows.