Skip to content

hey, i'm arjav jain

I build things,break things,and ship a lot.

building Torch, automation that writes dev docs from Figma changes

intern at ElevenX · CS/AI at Bennett '29

in short

I build products for the quiet problems people carry every day.

The method never changes: start from how someone actually behaves, then engineer around it.

I ran small businesses before I wrote serious software, so product thinking and psychology came first and the code arrived to serve them.

Most of what follows began as a hypothesis, shipped fast, measured honestly, and rebuilt without ceremony.

Featured Work

ChessMate

shipped · live at chess-mate.app

chess-mate.app

the problem

Chess engines tell you the best move, not why yours was worse. Evaluation without explanation makes you dependent on the engine instead of better at the game.

what I built

A web-based improvement platform that pairs Stockfish evaluation with an AI mentor chat: analyse your games, understand your mistakes, and train the weaknesses it finds.

what I learned

The engine was the easy part. Turning evaluation into explanation, something a club player can act on, is where the product actually lives.

ReactTypeScriptSupabaseStockfishTailwind CSS
Code, ChessMate
chess-mate.app

Mirror

shipped · live in production

usemirror.dev

the problem

Journaling apps are good at storing thoughts and bad at revealing patterns. Your own cognitive biases are, by definition, the ones you can't see, so writing alone rarely turns into self-awareness.

what I built

An AI-powered self-reflection journal that detects cognitive biases in what you write, tracks emotional patterns over time, and turns daily entries into genuine self-awareness.

what I learned

Building the bias detector taught me more about my own patterns than the journal did. Products that watch behaviour have to earn trust before they earn attention.

ReactSupabaseTailwind CSSSentiment AnalysisEdge Functions
Code, Mirror
usemirror.dev

Torch

in progress

the problem

Design changes travel to developers through meetings, screenshots and guesswork. The information already exists in Figma; the handoff is where it gets lost.

what I built

Monitors your Figma files, detects changes and generates developer-ready documentation automatically. No meetings, no manual spec writing, no back-and-forth. Currently in progress.

what I learned

Building on someone else's platform means designing inside their constraints. Half the work is understanding the Figma file model, not writing features.

TypeScriptFigma APIAutomation
torch (in the works)

Skills

what I build with

  • Python
  • TypeScript
  • React
  • Next.js
  • Supabase

what I think in

  • Systems
  • Products
  • AI
  • User behaviour
  • Experiments

Not everything deserves a case study. Some things were built to explore an idea, learn a tool, or answer a question that would not leave me alone. They still shipped.

the bench

  • exp-01 · CS50 final project

    Daily Digital Diary

    A Chrome extension that turns your browsing history into mindful insight.

  • exp-02 · Smart India Hackathon

    SwachSarthi

    Waste collection and compliance tracking for citizens and government bodies.

  • exp-03

    Stackit

    A realistic 3D LEGO-building experience in the browser.

Journey

six stops: from first venture to first CTO seat.

  1. 2024

    Marketing agency

    Co-founded DigiArc: brand identity, websites and marketplace listings for small businesses.

  2. 2024

    Clothing brand

    Launched Cult Notice, a Gen Z clothing label built from scratch.

  3. 2025

    Seasoning unit

    Set up a small seasoning production unit.

  4. 2025

    ElevenX internship

    Joined the sales and marketing team: leads, outreach and growth.

  5. 2025

    Makhana Express

    CMO & CTO for the import-export business.

  6. now

    CTO at Wroob

    Leading technology at Wroob.

  7. currently...

    • building random tools.
    • reading obscure essays.
    • watching people use software.
    • thinking about behaviour.

“Building is genuinely the only thing I'm good at. Everything else is just noise I'm trying to tune out.”

the honest version, straight from the README

Articles

  1. Why I Build a Tool I Am the Only User Of

    Mirror has one guaranteed user: me. That turns out to be the best product feedback loop I know.

    draft

  2. The Honest Chess Problem

    Chess engines are honest to a fault. Making that honesty useful is a design problem, not an engine problem.

    draft

  3. Technology Should Shape Better Behavior

    The best products do not just serve behaviour. They bend it somewhere better.

    draft

  4. Let the Robots Solve Utility

    Utility is a solved problem for machines. The interesting work starts after.

    draft

Contact

open to internships & founding-team roles

Good products start with people, not code.

Got an idea, a project, or just want to say hi?

I'm always up for a good conversation, a collaboration, a cool AI problem, a startup you're thinking about, or swapping ideas over coffee.

or find me here

usually reaching for Python · TypeScript · React · Next.js · Tailwind CSS · Supabase · Scikit-learn · Pandas · UX Research · Design Thinking