Comment by wasi0013

Experimenting with various AI models via GitHub Co-Pilot on an extremely niche project to see how far these models have progressed. Used like ~60% of the premium quota to develop the following projects:

website: https://murajah.pages.dev/

Play Store: https://play.google.com/store/apps/details?id=com.murajah.we...

Manual audio splitting tool for the above project: audio-splitter-6b3.pages.dev/

I've always been skeptical of AI-generated code. This is my first experiment with AI agents, where the full code base, implementation, debugging, and deployment are done using AI Agents MCPs.

Used VS Code all the way, i.e., all the source codes, including the code to generate the Google Play Store APK. I only reviewed the source code before committing and helped debug by suggesting ideas/algorithms.

Mostly used Claude Haiku 4.5 like 75% of the time, where it failed, switched to the sonnet 4.5 or GPT 5 codex. Interestingly, when debugging, sometimes one model struggled even after numerous iterations/feedback loops, but then the problem was solved instantly as soon as I switched to another model.

The source codes are available here:

https://github.com/wasi0013/Murajah

https://github.com/wasi0013/audio-splitter

Initially, I thought the audio splitting could be done automatically using some AI models from Hugging Face or Whisper. But the audio files have some complex repetitions; the output was miserable.

So, for now, this splitting is done manually using a Web UI (The audio splitting tool splits the large audio files into multiple small audio files, think of it like a long paragraph is split into multiple sentences.)

I will attempt again to automate this splitting task using AI, after drafting a game plan for tackling the challenges. I'm thinking of using energy drops and other similar factors to create segments.