Posts in
16 posts
Learn how to use Modal — a serverless GPU platform with $30/month free credits — to run data science and machine learning workloads from your local Jupyter notebooks and Python scripts. Covers setup with uv, notebook integration, GPU selection, and common pitfalls.
Last quarter, I needed to forecast SMS volume for budget planning. The catch? Our SMS volume directly depends on how many locations we operate—and that number is also growing. I couldn’t just extrapolate historical trends; I needed a model that accounts for this external...
A complete guide from dbt fundamentals to Python-powered transformations
A comprehensive quick reference for Cypher, the query language for graph databases. This guide focuses on Cypher syntax and patterns that work across Neo4j, Kuzu, and other graph databases supporting Cypher.
Execute Jupyter notebooks locally in VS Code while leveraging Databricks compute infrastructure. Your code runs on Databricks’ serverless compute, but you edit and manage files locally—giving you the best of both worlds: local development experience with cloud compute power.
Markmap is a simple yet powerful tool that transforms markdown files into interactive, visual mindmaps. It eliminates the need for complex mindmapping software—just write markdown, and Markmap handles the rest. The result is a beautiful, interactive HTML visualization...
A streamlined guide to setting up LaTeX on Ubuntu ARM servers and using it with VS Code
The COPY command in DuckDB and MotherDuck is a versatile tool for importing and exporting data. This guide provides a concise overview of how to use COPY both from the DuckDB CLI (SQL only) and from Python, including workflows with Ibis and pandas. Use this as a quick...
DuckLake is revolutionizing the lakehouse architecture by combining the simplicity of DuckDB with the power of modern data lake formats. In this comprehensive guide, I’ll walk you through building a complete DuckLake solution in two parts: first creating a local...
As a Data Science Masters student, I’m constantly working with mathematical concepts. From the calculus behind gradient descent to the linear algebra that powers PCA, math is the bedrock of everything we do. Recently, while tackling some homework, I stumbled upon a...
This guide provides step-by-step instructions to install a self-hosted QuickChart instance on an ARM-based server (like Oracle Cloud Ampere A1) running Ubuntu 24.04, using Docker Compose and Caddy.
OpenBB has evolved. While known for its powerful terminal and, previously, a self-hostable web application, the core of the self-hosted OpenBB Platform is now its robust set of API endpoints. This post will clarify what the current self-hosted OpenBB Platform offers, how...
Time Series is a topic that I’m very passionate about, because for many years I have worked in the preparation of Forecasts and Budgets for Revenue, Expenses and Headcount in Excel, but so far, I have never seen the use of Time Series techniques at work, to this day,...
I recently met with a group of colleagues from the accounting firm (KPMG), where I worked many years ago, and I found it hard to explain to them, what I do in the field of Data Science and what Artificial Intelligence is and is not, so after giving it a lot of thought, I...
This is a quick reference guide about Git so I can have it at hand. I took the data from the official documentation located at: https://git-scm.com/docs
Databricks is an analytics and data engineering platform that sits on top of Spark, an analytics engine for big data processing and machine learning. Spark uses in-memory processing using a distributed computer platform of clusters that work as if they were a single one.