“Philosopher’s Stone”. Created using Microsoft’s Designer (2023) in Copilot.

Today’s post is about building a virtual assistant and the lessons learned along the way. It can answer questions from my Obsidian notes, search the web, and execute Python code.

Picture by Dan Cristian Pădureț (Pexels)

In the ever-evolving landscape of software development, agile methodologies have gained significant popularity due to their flexible and collaborative nature. One such approach is Extreme Programming: Despite the funny name it is worth a second look, even more so than the omnipresent Scrum. Just as Scrum, XP is about simplicity, communication, feedback, respect, and courage. But in addition, you will see that XP fills some of Scrum’s gaps when it comes to the practical details of how to write code in an agile project. In this article, we will explore the key principles and practices of XP and how they contribute to the success of software projects.

“Cosmic Uncertainty” (this image was created with the assistance of Stable Diffusion)

When using machine learning (ML) applications, end-users often seek a reliable measure to assess the uncertainty of predictions. This is especially true when the stakes are high, such as medical diagnosis or accident prevention. One such measure are prediction intervals. When you’re working in python, they are basically just one more wrapper for your model. From my experience, the value of an uncertainty measure is often underestimated by end users at the start of the project. But in the end, it can make all the difference when it comes to user experience and adoption.

An ongoing collection of my origami models.

Have you ever tried to debug your Streamlit app like a normal python script? In case you want to see how it’s done in VS Code, continue reading this brief step-by-step guide.

Picture by Ron Lach (Pexels)

Data science projects can be complex and challenging to manage, especially when there are numerous stakeholders involved. The success of these and other software projects often hinges on getting the initial requirements right. A structured approach can help ensure that the project goals are clear to all participants and the development team has a common understanding of all the requirements.