01 /

About

in my own words

I am an AI engineer and full-stack developer, in my second year of a B.Tech in Computer Science (AI) at Vedam School of Technology, Pune. I build agents and RAG systems, ship the backend, web and mobile around them, and evaluate LLMs for a living.

I started college solving LeetCode problems. Four months later I was building multi-agent systems at hackathons, and by the summer I was the only engineer on a client's backend, PWA and app. Right now I do three things at once: full-stack intern at Third Shade Media, certified LLM evaluation expert at Deccan AI Experts, and co-founder of CampusCritique.

I was born in Jamshedpur, my family is from Bihar, I grew up and went to school in Delhi, and I study and work in Pune. School also left me with some German, a Scouts badge (Dwitiya Sopan) and a year as a prefect. I watch a lot of anime; Bleach: Thousand-Year Blood War is the current one.

The things I care about are simple: build for a specific person, measure what shipped, and say plainly what did not work. Every number on this site links to its source, and the failures get the same space as the wins.

02 /

How I work

from the projects, not a trait list
  • Every number has a source. A metric without a query, run file or command behind it does not go on the page.
  • Not every problem needs an LLM. Rules and scoring first, the model where judgement is needed.
  • Make failures visible. Failed jobs, collapsed models and bad webhooks are part of the story, so they are on the home page.
  • Hold unknown facts as null. Never guess in production content; fail the build instead.
  • Write it down. A structured note after every milestone, which is why the changelog exists.
  • Blocked means find another route. Noesis was won through a mid-event twist and a dead connection.
03 /

Engineering decisions

the why behind the case studies
03.01

Why a queue between ingest and delivery?

Notify commits the event first, then publishes to RabbitMQ, so a provider outage never blocks the caller and every job can be retried or recovered from the database.

03.02

Why rules before the LLM?

KisanMind filters and scores with rules first and lets the model rerank the shortlist. Deterministic where it can be, testable, and cheaper to run for a farmer on a phone.

03.03

Why focal loss?

OilTrace runs on patches where oil pixels are rare; Dice loss kept collapsing to the empty mask. Focal loss concentrates the gradient on the pixels the model gets wrong.

03.04

Why idempotent webhooks?

Payment gateways retry, reorder, and send refunds through the same pipe. CampusCritique keys on the gateway event id and guards refund updates.

03.05

Why build gates?

Margo Rubber had contradictory source material. A build that fails on a missing fact is cheaper than an apology to a customer.

03.06

Why does this site refuse questions?

Ask this site answers only from retrieved chunks above a calibrated similarity threshold and says so when nothing qualifies. A wrong answer about me costs more than no answer.

04 /

Timeline

2025 to present
  1. Oct 2025First hackathon, second prize (cultural-club)
  2. Mar 2026First place, Noesis Hackathon (AskMyNotes)
  3. Apr 2026Co-founded CampusCritique; top 2 at HackWarts (KisanMind)
  4. Jun 2026Third Shade Media internship; Deccan AI Experts certification
  5. Aug 2026OilTrace wins the Smart India Hackathon internal round
  6. Sep 2026Notify in production for two tenants; Humraah in store review
05 /

Elsewhere

8 projects on Work. Every one is sourced; the ones marked measured have an evaluation table.

keysKeyboard
j / k
next / previous row
Enter
open the focused row
⌘K or /
search, or ask this site
g then h w a c
go home, work, about, contact
t
toggle light and dark
Esc
close the palette or this map
?
this map