AI Money Labs
Premium

Cases

An AI quiz generator built alongside a day job — QuestGen

BMostly confirmedLevel 2

A data scientist built an AI quiz and question generator while keeping a day job. We do not publish a revenue figure: the only source is the founder. What we do publish is what was sold, to whom, and why this case is hard for a non-specialist to repeat.

Founder
Ramsri Goutham Golla
Country
IN
Team
—
Startup cost
—
To first revenue
—
Automation
high
Reported revenue
Revenue not disclosed

What is sold

SaaS subscription, free to start

Pricing sits behind signup rather than on the homepage, so we could not confirm the figures.

Who buys

Teachers and schools, HR teams, publishers and edtech

AI's part and the person's part

Already had

Data scientist. This is the most important field in this case — natural-language expertise maps directly onto the product's core feature.

If this is large, the AI is not the reason

What the AI did

Feed it a document, image, video, or web page and it produces multiple-choice, true/false, and fill-in-the-blank questions. Generating the questions is the product, and it replaces work a person used to do by hand.

the part a tool replaced

What the person did

The founder builds and runs it. As checked on 2026-08-14 the site shows no careers page, no team page, and no investors, and the footer links the founder's own email address. The same founder also runs another AI app (AiArtist).

stories that omit this do not replicate

The three columns are not proportional to contribution — we did not measure that. This is a schematic for separating them.

How customers were found

  1. Product-led — free trial, no card required

  2. Many feature-specific landing pages (per question type and input format)

  3. The founder's public presence

  4. Paying customers

    Pricing sits behind signup rather than on the homepage, so we could not confirm the figures.

Channels we could confirm from public sources. Conversion rates were not disclosed, so they are not shown.

Flow: Product-led — free trial, no card required, then Many feature-specific landing pages (per question type and input format), then The founder's public presence, then Paying customers

How we verified this

Evidence strengthB· 1 source(s)

Public dashboard or platform figure. May be self-reported

We confirmed on 2026-08-14 that the product resolves and checked its operating scale directly: no careers, team, or investor pages, and a founder email in the footer. The founder is named. But the only source for revenue is the founder, with no payment-processor link or third-party confirmation, so this is a B. The site's “150,000+ users” and “2.5M+ questions” are self-reported; we did not verify them and do not rely on them.

See the grading rules

What made it work

What the AI removed was the labour of writing questions — and the market chosen was one where that labour was already expensive (teaching and assessment). The decisive choice looks less like picking a good tool and more like picking a place where the disappearing labour was costly.

Risks

⚠️ The main risk is replicability. The founder is a data scientist and that expertise overlaps the product's core feature. Separate “because of AI” from “already knew this field” and this case leans heavily on the second. A non-specialist should not expect the same result. General-purpose models also keep absorbing this feature, which is continuous substitution pressure.

Does this work in Korea?

The structure transfers: nothing about it is US- or India-specific. Selling subscriptions to individuals abroad does require sorting out payment and sales tax (which differs depending on whether you use a Merchant of Record). Selling to Korean schools and public institutions adds a separate procurement process.

Replicability in Korea2/5
Difficulty3/5

Solo-feasible: yes · Automation: high

7-day action plan

  1. Step 1. Write down a field you already know

    this case started with domain knowledge, not with a tool

  2. Step 2. In that field, pick one artefact people produce repeatedly by hand

  3. Step 3. Make one of them with AI and record how long it took

30-day action plan

  1. Step 1. Show it to five people who actually use that artefact and record their reaction

    do not sell

  2. Step 2. Publish it in a form people can use for free and watch whether they do

  3. Step 3. Attach a price only once people are using it. If nobody does, change the artefact, not the price

Run the numbers

We do not print a single projected figure here. Put your own numbers in instead.

Break-even calculator →