Welcome to my blog! This is where I'll be sharing my thoughts on software engineering, AI, ML modelling, fine tuneing models and building products that actually solve problems. I've been meaning to start writing for a while, and here we are.
Writing forces clarity. When you try to explain something you've built or a decision you've made, you quickly discover the gaps in your own understanding. That's the whole point.
If you can't explain it simply, you don't understand it well enough.
— Richard Feynman
I am based in Christchurch, New Zealand and moved here from Bhutan in my late teens after finishing highschool. Ethnicity wise, I am nepalese. I learnt how to design web and mobile apps and learnt to code myself by watching youtube tutorials and reading books while working as a structural engineer (which I have a bachelors degree in) for around 5 years. Got my Masters degree in Computing to validate my skills and achieved that with distinction.
My approach to most things is just face the problem first. Try it. Break it. See what happens. I've wasted more time planning around problems I hadn't actually hit yet than I ever have on the problems themselves. The mitigation strategies, the what-ifs, the "let me think about every edge case before I start" mindset. All of that costs more time than just running into the wall and figuring out which part of it is actually solid.
And the things that look intimidating at first? They're usually just big, not hard. Break them down and they stop being scary. When I learn something new I start with the basics, experiment, move the dial up slowly. If I get stuck I pretend I'm teaching it to someone younger. Imagining myself as both the teacher and a younger version of myself trying to understand makes things way easier to digest. If I can't explain it simply to that imaginary student, I don't understand it well enough yet.
I also don't like big words. If a simpler word does the job, use the simpler word. Overcomplicating things doesn't make stuff sound smart, it just makes these things harder to understand.
I'll be writing about the things I work on from building voice AI platforms to experimenting with agentic systems and RAG pipelines plus many many more things related to AI. Stay tuned.