Python

In our previous post, we introduced function calling and learned how to do it with OpenAI’s LLMs. In this post, we’ll call the same cactify_name function from that post using Meta’s Llama 3.2 model, installed locally using Ollama. The techniques in this post should also work with other Ollama models that support function-calling.

In our previous post, we explored how to send text to an LLM and receive a text response in return. That is useful for chatbots, but often we need to integrate LLMs with other systems. We may want the model to query a database, call an external API, or perform calculations.

At Caktus, many of our projects use pip-tools for dependency management. Following Tobias’ post How to Migrate your Python & Django Projects to uv, we were looking to migrate other projects to uv, but the path seemed less clear with existing pip-tools setups. Our requirements are often spread across multiple files, like this:

Editor’s note: This post was updated on July 17th, 2025, to use the PATH_add hook for direnv and to recommend including uv sync --locked in your .envrc file.

Upgrading projects is somewhat equivalent to flossing, you know you have to do it, but rarely make time for it. After all, if the project is in active development, there are exciting new features to build. And we all know that new features > project upgrades. Well not to worry, Caktus wants to make you aware of some tools that will save you from considerable repetitive work & time while simultaneously modernizing your codebase. Combined, these tools will automate part of the upgrade process, decreasing the likelihood of neglecting parts of the codebase.

We use factory_boy for bootstrapping test data on many Python and Django projects at Caktus. Recently, we encountered a deprecation warning on an older project that had been using factory_boy for some time:

Recently, Caktus has been using Tailscale to manage VPN connections between Android tablets and a central server. We wanted to report on the devices connected to the network using the Tailscale API. While we could use tools like Celery to fetch data from the API and load it into a database—given its widespread use in the Django ecosystem—we also wanted to explore other options.

We regularly use tools like PostgreSQL, Pandas, and Jupyter Notebooks to analyze data here at Caktus. Recently, we were reviewing North Carolina traffic stop data for the NC CopWatch project and had the opportunity to use PostgreSQL's window functions, which are helpful when aggregating data.

Lately, there has been a lot of talk about scoring in the NBA because LeBron James surpassed Kareem Abdul-Jabbar with 38,390 career points. I have noticed that there is not much discussion about post-season scoring, so I searched for this dataset on Kaggle (nba_playoffs.csv) which contains the top 25 all-time post-season scoring leaders. Post-season scoring is its own beast. Since teams face one opponent multiple times in a row, they can better concentrate on the opposing team and its individual players, particularly star players. This results in improved defenses across the board. However, the post-season also means players improving their game. What is the result of improved defenses and players alike? Only elite players score consistently and thus, only the NBA's elite are on this list. This post will first examine the dataset using Pandas and then use Seaborn to graph such data.

Wouldn’t it be useful if we could document changes in our life and revisit them later at will? It would allow us to better analyze situations, remember what we were thinking, or help us remember how we got to our current state. Although no such tool currently exists for changes in life, one such tool does exist in Django. It is called django-simple-history.