It has been a while since I have posted anything here. When I first started this site, it was mostly a small portfolio/blog while I was working through my computer science degree. A lot has changed since then. I finished my B.S. in Computer Science, completed the IBM Data Engineering certificate, and I have been working as a Programmer Analyst at Stephen F. Austin State University. So I figured it was time to wake the site back up. I do not want this to just be an old student portfolio sitting online forever. I still want it to be a place where I post projects and technical notes, but I want it to better reflect what I am working on now. Lately that has been a mix of homelab work, Python projects, automation, data cleanup, reporting workflows, and some AI experiments that I think are worth documenting.
The Homelab
My homelab has become one of the main places where I learn and test things. It is not some perfect enterprise setup, and I am not going to pretend it is. It is a place where I can try ideas, break things, rebuild them, and slowly make my own workflows better. I use it for self-hosting, development environments, storage, automation experiments, and whatever project I happen to be working on at the time. One thing I have been thinking about more lately is making my projects easier to come back to later. It is easy to throw together a script or app and then forget how it works six months later. So I have been trying to get better about source control, dev containers, testing, documentation, and repeatable setup steps. That is not always the most exciting part of a project, but it matters. If I can rebuild it, test it, and understand it later, then it is a lot more useful.
Glitch and Deskling
One of the projects I am most interested in right now is something I am calling Deskling. The companion itself is named Glitch. The basic idea is to build a small desktop companion in Python. I do not want it to just be a chatbot window. I want it to feel more like a small simulated creature that lives on the desktop. Something with memory, routines, a simple personality, idle behavior, and eventually maybe voice, camera input, and other ways to interact. This is still an experiment, and I am trying to be realistic about it. The first version does not need to be impressive. It just needs to be a real starting point. Some of the things I want to explore with Glitch are:
- How should a small AI companion remember things?
- What should be short-term memory vs. long-term memory?
- How do you keep personality consistent over time?
- How do you test changes without breaking the whole thing?
- What should the companion do when I am not actively talking to it?
- How can it feel like a creature instead of just another assistant? I also have some longer-term ideas around giving Glitch temporary ways to interact with the world. For example, I have a small Tello EDU drone that could eventually act as a kind of scout or vehicle. But I do not think of that as Glitch’s real body. It would just be one possible tool or shell. The actual companion should be separate from whatever interface it is using. That is the kind of thing I want to write about more here. Not as if I have all the answers, but as a record of the experiments.
Automation and Workflow Experiments
I have also been spending a lot more time thinking about practical automation. A lot of useful programming is not flashy. Sometimes it is cleaning up a spreadsheet, validating a file, converting data from one format to another, adding better logging, or making a process easier to repeat. Those things can seem small, but they can save time and prevent mistakes. That kind of work has changed how I think about projects. I am less interested in building random apps just to have more GitHub repos, and more interested in building things that solve an actual problem or teach me something useful. Some of the areas I want to keep working on are:
- data cleanup and validation
- small internal tools
- repeatable scripts
- better logging
- testing workflows
- dev containers
- source control habits
- documentation that future-me can actually understand That last one is important. Future-me is basically another user, and he does not remember anything.
What This Site Is Going To Be
Going forward, I want this site to be more of a technical notebook and project journal. Some posts might be polished. Some might just be notes about what I tried, what broke, and what I learned. I want to write about homelab projects, Python experiments, Glitch development, automation, data workflows, retro computing, and whatever else I am learning. The older posts are staying because they are part of the path. They show where I was when I started this site. But I want the site to reflect where I am now too. So that is the update. The site has been quiet, but the lab has not been.