Let’s talk

Most businesses know AI can help. I build the part that actually works.

Real products, shipped.
Here is how.

Kaavya GuptaScroll to exploreAhmedabad, India

I study commerce, not computer science. I learned to build by shipping real products for real businesses .

Selected work

Software, AI and websites, built for real businesses and used every day.

1 of 4, Inventory planning software

&circus

Turns demand for every SKU into what to restock and how much fabric to buy.

  • Next.js
  • React
  • TypeScript
  • Tailwind
The &circus planning dashboard, rebuilt with demo data
3 of 4, Stock research platform

BroMoney.ai

Every stock scored, simplified.

  • APIs
  • Cloudflare Workers
  • Caching
  • Custom DNS and deployment
BroMoney.ai home page
4 of 4, Website for the Bloomberg Lab

Lab of Leaders

The site for the Bloomberg Lab at CHRIST University, powered by Career Pandit.

  • About 10 pages
labofleaders.in home page

What I build

Four kinds of work, each one proven on a real project. Hover a row to see it.

Small tools, real use

Automations and open source tools, built for a firm, for other builders and for myself.

Outreach skillRunning
  1. Checking Apollo credits
  2. Finding verified leads
  3. Building the Excel sheet
  4. Writing a Gmail draft for each lead
  5. Drafts ready in Gmail
XLSleads.xlsx
A free, open source Claude skill

Outreach skill

Finds verified leads on Apollo, checks credits first, builds an Excel sheet and writes a personalised Gmail draft for each lead.

View on GitHub
A crab that watches Claude Code

Clawd

Walks above the Mac dock and mirrors live Claude Code sessions. Speeds up while Claude works, sleeps when idle, and stops with a ! when a session is waiting on you.

View on GitHub
Automations for a chartered accountancy firm

Shubham Maloo & Co.

One-click n8n workflows that write content ideas and scripts into Google Sheets and keep project reports updated in Google Drive, using AI agents, SerpAPI and Google Workspace.

How it works

From the first call to a product your team uses every day.

  1. 1

    Talk

    You tell me the problem. We work out what a good result looks like.

  2. 2

    Plan

    I map out what to build, and which parts AI should and shouldn't handle.

  3. 3

    Build

    I build in short rounds and show you working versions early.

  4. 4

    Launch

    It goes live, and we keep improving it from real use.