Problem Solving in My Home Lab by Rory
Rory's entry into Varsity Tutor's November 2026 scholarship contest
- Rank: 97
- 0 Votes
Problem Solving in My Home Lab by Rory - November 2026 Scholarship Essay
The academic skill I want to develop further is problem solving, and honestly it is the skill that got me into IT in the first place. I like figuring out why something is broken and then actually fixing it, not just reading about it. School teaches the concepts, but the skill only gets better when you use it on a real problem where nobody is handing you the answer.
The project that taught me the most about this was in my home lab. I wanted to run an AI model on my own computer, with no internet, no cloud, nothing phoning home. Just my desktop and my own hardware. I had been using free tiers of Google Cloud and Azure for school projects, and everything ran somewhere else. I wanted to know what was actually happening under the hood, so I made it a project. I installed a graphics card in my desktop, set up a virtual machine to keep things clean, and downloaded Llama, an open model I could run locally. Nothing about it went smoothly at first. The model files were huge. My first attempt barely ran because I had not given the machine enough memory. I spent a few nights reading documentation, trying different settings, and watching my system crawl until I figured out what each setting actually controlled.
That is where I actually learned the most, you know, because in class AI was always this abstract thing. You send text to an API and an answer comes back. But when the model is running on your own machine, there is no magic. You see exactly how much memory it takes to hold the weights. You see how the speed of your answers depends on your hardware, not on some server farm. When I changed the context length, I could watch the memory usage climb in real time. When I picked a smaller model, the answers got worse in specific ways I could actually measure. Every time something broke, I had to narrow it down step by step until I found the real cause, and that whole process is the skill.
It also changed how I think about privacy and infrastructure. A model running on my desk does not send my data anywhere. For a field tech who has spent years running cable through ceilings and swapping out access points, that clicked. The cloud is just somebody else's computer, and sometimes the right answer is your own computer.
I want to keep getting better at this because it is pretty much what my whole future is going to need. I am studying Information Systems Technology at Seminole State College and I want to work in cloud infrastructure, eventually as a solutions architect or a solutions engineer. In that kind of work, the problems are never the textbook kind. Systems break in weird ways, and the person who can stay calm and just break the problem into pieces and test one thing at a time instead of guessing is the person who gets it fixed. That is what I am practicing now, in my lab and at my job, and it is the skill I want to keep building all the way into my career.