Podcast
I host Positively Yours
33 episodes about healing, change and the parts of growth nobody posts about. It is the least technical thing I do and probably the most honest. Season 5 is live, with a new episode every week.
Listen to the podcast →About
Platforms · Messaging · AI
Hi, I'm Kashish. I build software for a living and think about people for a hobby, which turns out to be the same job.
I'm an engineer with nine years in enterprise software - messaging platforms, multi-tenant SaaS, and the interfaces people run them from. Most of that time has gone into one thing: taking systems that are genuinely complicated and making them feel simple from the outside.
I started in frontend, which is where I picked up the habit of caring how something feels to use and not only whether it works. The work widened from there into platform architecture and real-time messaging infrastructure, and for the last three years into AI - mostly the evaluation, cost and failure handling that decide whether a language model is a product or a demo.
I hold an MBA in Operations and Data Science. It confirmed something I already suspected: the engineers who do well next are the ones who understand the product and the people as well as the stack.
Away from the screen, I hail from the mountains - which is probably why I still measure a good day by how quiet it was.
Off the clock
None of that is a personality section. It is where most of my better product instincts have actually come from.

Practice
Concept through execution
I build digital products end to end, from the first sketch to the thing running in production - with designers, product managers and other engineers the whole way.
The barAnything that only works in the demo is a prototype wearing a suit.
The arc
Frontend to platforms to AI
Start
Frontend
Where the habit came from: caring how something feels to use, not only whether it works.
Then
Platform architecture and real-time messaging
Multi-tenant SaaS, and conversational messaging across RCS, WhatsApp and SMS. Systems where the hard part is never the happy path.
Last three years
AI in production
Mostly the evaluation, cost and failure handling that decide whether a language model is a product or a demo.
Alongside
MBA, Operations and Data Science
It confirmed a suspicion: the engineers who do well next understand the product and the people as well as the stack.
Beyond engineering
Four standing interests
Technology is a big part of my identity, not the whole of it. None of the below is a hobby I do to look interesting. It feeds straight back into how I design things, mostly by reminding me that software is for people who are tired, busy and half distracted.
Why people do what they do. This is the part of product design that never fits into a framework.
Attention is the scarcest thing I have, at work and everywhere else. So I practise it on purpose.
New places reset my defaults. A few of my better product instincts showed up nowhere near a desk.
The same learning habit I apply to the stack, turned on myself. It is slower and less comfortable.
Also
Podcast
33 episodes about healing, change and the parts of growth nobody posts about. It is the least technical thing I do and probably the most honest. Season 5 is live, with a new episode every week.
Listen to the podcast →Build with intention. Keep learning. Leave every product better than you found it.
How I try to work