*
i'm hrushikesh — an ai developer at te connectivity in bengaluru. i spent 3.5 years in .net and react, then made the jump to building llm-powered systems. i like shipping things that work in production, not just in demos.
* first principles
build, don't theorize
i learn by making things work. tutorials are a starting point, not the destination. if i can't run it locally, i don't understand it yet.
compound over time
3.5 years of .net gave me production instincts that most ai devs skip. debugging distributed systems, writing apis that handle real traffic, shipping code that other teams depend on. every skill compounds — nothing is wasted.
read widely, build deeply
i read google research papers and stoic philosophy in the same week. breadth in input, depth in output. the best ideas come from connecting things that don't usually go together.
share what you learn
writing about search algorithms forced me to actually understand them. contributing to open source taught me more than any course. if you can't explain it simply, you don't know it well enough.
* books
i read a mixed variety — ai papers one day, stoic philosophy the next. here are the ones that shaped how i think and work.
on building
- introduction to agents — alan blount, antonio gulli et al. (google)
- agentic design patterns — antonio gulli
on thinking
- the almanack of naval ravikant — eric jorgenson
- atomic habits — james clear
- how to stop worrying and start living — dale carnegie
on grit
- can't hurt me — david goggins
- discipline is destiny — ryan holiday
on people
- how to win friends and influence people — dale carnegie
- tuesdays with morrie — mitch albom
- half torn hearts — novoneel chakraborty
* currently exploring
- -tweets & blogs from sam altman, paul graham, andrej karpathy
- -arxiv papers on multi-agent systems and data science automation
- -whatever shows up in my feed that makes me think differently