Cognitive Monitoring System
An ML platform that reads keystrokes, mouse drift, and idle patterns to predict stress in real time.
I'm a final-year CS student from Jaipur who likes to live near the seam — between software and the people who use it, between the cloud and the bit. I build things that work in production, then I write about why they work. I read long-form, I play cricket on weekends, and I'm suspicious of code that doesn't have a user in mind.
Two case studies I keep coming back to. Each one taught me something I now refuse to ship without.
An ML platform that reads keystrokes, mouse drift, and idle patterns to predict stress in real time.
A full-stack marketplace for student accommodation — listings, bookings, payments, in production.
Three internships, each in a different part of the stack — IoT at the edge, ML in the middle, and DevOps at the cloud.
> Deploying a Java application on Azure using DevOps
Designed, containerised, and deployed a Spring Boot application on Microsoft Azure. Built an end-to-end CI/CD pipeline with GitHub Actions, containerised with Docker, and shipped to Azure Web App / AKS. Added monitoring, logging, and security baselines so deployments were actually reliable — not just fast.
> Digital Number Prediction from Stock Images
Built a computer-vision model that recognises digits in stock images — a small but useful grounding exercise in dataset curation, preprocessing, and evaluation. Trained on labelled image data and shipped a working inference pipeline.
> Green Field: Smart Irrigation Control System
Designed a smart irrigation system that combined soil-moisture sensors with weather APIs to automate watering schedules. The estimate: ~40% less water usage and better crop-yield simulation. The real lesson: hardware, software, and environment all have to agree.
Seven processes running constantly. They keep me a person, not a job title.
> Now playing — R.D. Burman · 'Dum Maro Dum'
I keep a queue of pre-2000 Hindi film scores and let them run while I work. Slow, melodic, and never competes for attention. I don't really listen to lyrics — I listen to arrangements.
> Re-watching — 2005 Ashes, the Overton spell
I play on weekends and watch more than I should. Test matches over T20 — five days is the point. I've started tracking my own batting in a notebook, which is the kind of nerd move I won't apologise for.
> Just finished — Outer Wilds. Crying, still.
I don't really play competitive games. I play things that make me think differently about systems — Outer Wilds, Disco Elysium, FTL. I keep a notes file of every trick a game teaches me about engineering.
> Next — 14h to Jaisalmer, winter desert run
I take the train everywhere I can. There's a stretch of the Jaipur–Jodhpur line at golden hour that I'd happily ride monthly. I keep a list of small stations with good chai.
> Cup 3 of chai, made from the steel filter
I do pour-over on weekends with a Hario V60, but most days it's masala chai from a steel filter. The ritual matters more than the caffeine. I have a small notebook where I rate each batch.
> Reading — 'The Beginning of Infinity' · David Deutsch
I'm suspicious of books that promise to optimise me. I read long-form essays and science — Deutsch, Taleb, early internet writing from the 90s. I keep a public notes file of passages I want to come back to.
> Building — cognitive load, visible to the user
Lately I've been obsessed with software that tells you what it's doing while it's doing it. Visible state, honest error messages, dashboards a non-engineer can read. Less 'magic', more 'this is what's happening'.
The tools I keep coming back to. Ordered by how often I actually reach for them — not by how good they look on a CV.