NEW LLM Engineering course is live →

We build software shops run on. Then we teach it.

A software company in Bhilwara. Our billing software runs three jewellery shops every working day — and the courses are written from building it. Every lesson has a video in English and Hindi.

Explore courses Start a project

3Shops use it daily

7Courses

88Video lessons, Hindi + English

2Free courses

bill.js

// what a jewellery bill actually needs — from the software
// three shops in Bhilwara use every working day
export function billLine({ grams, ratePerGram, makingPct, gstPct = 3 }) {
  const metal  = grams * ratePerGram;
  const making = metal * (makingPct / 100);
  const gst    = (metal + making) * (gstPct / 100);

  // round once, at the end — never on each line, or the
  // printed total stops matching the arithmetic
  return { metal, making, gst, total: Math.round(metal + making + gst) };
}

// → shared on WhatsApp, saved against the customer's khata
        

Why we are different

Most course platforms have never shipped the thing they teach.

We run an engineering studio first. Every lesson on this site traces back to a project that went to production, had bugs, and got fixed at an inconvenient hour.

Built the way we build for clients

No framework, no build step, no dependencies. The assistant in the corner runs TF-IDF retrieval we wrote from scratch. The AI Lab runs TextRank, naive Bayes and a sentiment lexicon — all in your browser, all offline. Open DevTools and read it.

0 dependencies Offline AI 100 Accessibility

Try the AI Lab

Shipped, not theorised

Our billing software runs three jewellery shops in Rajasthan — bills, stock and who owes what, every working day.

Two free courses

Complete, not trials. 58 hours of web and ML, no card required.

We say no, and we say why

We have talked clients out of rebuilds, out of microservices, and out of RAG when keyword search would have done. The discovery week exists so you find out before you have paid for the wrong thing.

Weekly deploys

A staging URL from day one. You watch it get built instead of waiting for a reveal.

Live, on this page

Not a video. Not a screenshot.

Paste any text below and watch a TextRank summariser build a sentence similarity graph, run PageRank over it, and pull out the sentences that carry the document. It is the same algorithm taught in the ML track, running in your browser right now.

All six tools

Your text

Sentences

Learn