I measure what I build.
An ML engineer who costs systems before building them and instruments them before shipping. Most of my work starts with someone unable to explain what their system just did.
that's me, Shreyan ✦
How I got here
I started in applied machine learning and kept ending up on the same problem: systems that performed well in a notebook and behaved unpredictably once real users arrived. The interesting work turned out to be in the gap between those two states.
That led to costing systems layer by layer before writing code, and to instrumenting them so that behaviour and spend are visible from the first commit. It also led to writing, because the fastest way to understand something is to run the experiment and publish what happened.
Alongside client work I founded the Claude Developers India community, publish technical articles for Weights & Biases, and write on responsible AI policy, including a certification framework sent to MeitY and NASSCOM.
builders in Claude Developers India, founded and run by me
technical articles, each from an experiment I ran myself
tests across cc-habits, twelve of them security suites
What I believe about this work
A model chosen without its per-unit cost is a bill you have not read yet. The stack should be picked against a cost model, not against a leaderboard.
If the data cannot tell you whether the fix worked, the fix is a guess. Sometimes the honest deliverable is a diagnostic rather than a build.
A result that shows an approach does not work is worth as much as one that shows it does, and it is cheaper to learn early.
Easiest way to reach me
Three questions about your system and I will tell you whether it is worth a call.