JRVS: My First Open Source Project

I released my first open source project on GitHub and I am absolutely ecstatic to see what it grows into. 6 to 8 months ago I started to get into the world of local AI. I discovered technology that allowed you to run LLM models that are similar to the commercial ones, at different scales. This was the beginning of my Coding Journey.

The Problem I Discovered

When I got very used to using Ollama which allowed me to interact with LLMs that I downloaded, I realized one problem. That problem being if you wanted to use a model for a real conversation, at a certain point the model would time out. The timing out not only was annoying, but it halted the workflow. If I wanted to keep talking to the model about that topic I was out of luck.

Finding My Solution

This prompted me to look for an alternative interface for interacting with my local AI models. This is what started the rabbit hole for me. If I did not really feel the need to create the tool I have created for myself, I would not have had a chance to create software that the rest of the world can enjoy.

What I Learned

What I learned through creating JRVS is that the first step in creating software people actually want to use: Find a real contention people have with existing solutions/tools etc. Why invent a new wheel when I can improve the slightly broken one instead?

I think doing what I did benefits the open source community more, simply because I would be elated to find that my solution can solve the same issue for other devs, students and anyone interested. I think it’s more the energy behind what I am doing guiding every action that is helping me do one of the most exciting things ever: Contribute to a community that started me off.

Giving Back to Open Source

I am a Linux enthusiast. Half the software I use is open source. Just through math I need to contribute a little so I can say I didn’t just leech off the great open source community even though that’s just fine!

What I really learned through creating my first open source tool is that even if you love tech, the best outcomes in the industry will come from the unique intersection of people and tech.