Aiolum Submit a story
All stories
AI teaching assistant for lesson planning and AI-written-text detection · USA

Tabitha.ai

Solo founderExitB2CEducation
Outcome
Sold for US$4.25K, July 2023
Launch
March 2023
Team
1 founder, part time, a couple of hours a day; no employees
Outside funding
None reported; built part time on no-code tools
Main channel
Not stated
Reported, not verified

What is not verified

The headline figure, a sale for US$4,250, rests on one source: the founder's own interview in the No-Code Exits newsletter, published on 6 July 2023. The same interview is the only source for the US$38 a month of revenue, the 2 paying users, the US$19 price and the 9.5% marketplace fee. The buyer is not named, the Tiny Acquisitions listing is archived only as a script-rendered page with no readable text, and no payment-processor profile exists. A reader could look for the buyer's own account of the purchase, a Tiny Acquisitions record of the closed listing, or a later statement by the founder that repeats the figures. The headline number is grade D: one claim or one estimate, with no second source. Aiolum shows it because the story is useful, and marks it so that it is not mistaken for a verified figure.

Tabitha.ai was an AI teaching assistant that ran in a browser tab. It started as an AI detector, a tool that estimates whether a text was written by a language model such as ChatGPT, for teachers who suspect that an essay was not written by the student. The founder then added generative AI features, that is, features where a language model writes new text on request: lesson plans, assessments, topic research and draft feedback. The whole app was built with no-code tools, visual builders such as Bubble that let a person assemble a web application from blocks and connect it to outside services through their APIs, the interfaces that programs use to call other programs, without writing program code. The founder said in July 2023 that teachers paid US$19 a month for it as a SaaS subscription, software sold as a monthly service.

Milly Barker-DeStefano, an international lawyer from England who lives in Connecticut and runs a small venture studio beside a job as chief growth officer at a fintech startup, told the No-Code Exits newsletter in July 2023 that she built the first version, only the detector, in about two weeks of a couple of hours a day, without telling anyone, because she was not sure she could make the AI integrations work in Bubble. She said that she then listed the app on Tiny Acquisitions, a marketplace for very small online businesses, got interest the same day, and sold it for US$4,250, of which the marketplace took a 9.5% fee. At the time of the sale, she said, the app had about 100 free users, 2 paying users, revenue of US$38 a month and about 600 visitors a month. The buyer, she said, wanted to learn Bubble and had contacts among teachers.

Every figure on this page comes from that one interview, published on 6 July 2023, so the evidence grade is D, reported and not verified: no buyer statement, marketplace record or payment-processor profile confirms the price. The page also lists earlier tools for teachers in the Loot Drop database that died when free platforms from Google and others covered the same jobs. The domain tabitha.ai still sells a teacher assistant in October 2026 under a later owner.

Tabitha.ai: Sold for US$4.25K, July 2023. AI teaching assistant for lesson planning and AI-written-text detection, USA, solo founder, evidence grade D.
The story in one card. Share it with the address of this page.
01

The numbers, graded

Each row shows the evidence grade and the source numbers from the list at the end of this page. The grades are explained in How Aiolum verifies.

A no-code app with 2 paying users that its founder said she sold for US$4,250 in mid-2023, about four months after the first archived copy of its page; one source for every number
FactValueGradeHow we knowSources
Sale price, by July 2023US$4,250DThe founder's statement in the No-Code Exits interview published 6 July 2023. She said Tiny Acquisitions took a 9.5% platform fee. No buyer or marketplace record was found.[1]
Net to the founder after the marketplace feeabout US$3,846calcCalculated: US$4,250 minus 9.5%. The founder did not state the net amount.[1]
Revenue per month at the saleUS$38DFounder's statement, July 2023. Two paying users at US$19 a month.[1]
Price of the sale against monthly revenueabout 112 months of revenuecalcCalculated: US$4,250 divided by US$38. The buyer paid for the built app, not for the revenue.[1]
Users at the saleabout 100 free, 2 paying, about 600 visitors a monthDFounder's statement, July 2023. Elsewhere in the same interview she says 40 to 50 sign-ups before the sale; the two figures are not reconciled.[1]
Price per month, 2023US$19DFounder's statement, July 2023. The archived pages are built by script and show no prices.[1, 3]
Cost to serve one free-trial userabout US$0.25DFounder's statement, July 2023, the cost of the AI calls for a trial.[1]
Time to build the first versionabout two weeks, a couple of hours a dayDFounder's statement, July 2023. Her first app built alone in Bubble, including sign-up, log-in and payments.[1]
Team1 person, part timeDFounder's statement, July 2023: she built it alone beside a full-time job and did not want to build a team.[1]
Outside fundingnone reportedDNo investment is mentioned in either interview; the founder describes self-built side projects sold to other entrepreneurs.[1, 2]
Earliest archived copy of tabitha.ai26 March 2023CThe Wayback Machine index lists a first capture on 5 March 2023 with no stored page, then a stored page on 26 March 2023 that describes an AI teaching assistant. The link-preview image in that page carries a Bubble upload stamp that converts to 27 January 2023.[3]
The founder's earlier exitShout About, sold for US$15,000 in 2022CHer November 2022 interview with the same newsletter: a media-contacts app built with a co-founder on Bubble, with US$1,003 of monthly recurring revenue, listed at US$10,000 and sold for US$15,000.[2]
Months from the first archived page to the sale4calcCalculated: March 2023 to July 2023, the month the interview reporting the sale was published. The exact sale date is not stated.[1, 3]

How to read this table: a grade D value is a single founder statement in one interview, with no second source. A grade C value is read directly from an archived or current page. A "calc" value was computed on this page from the rows above it. No value is grade A or B.

The 2023 price as the founder stated it in July 2023 (grade D), and the prices on tabitha.ai in October 2026, which belong to a later owner (grade C)
PlanPer month, billed yearly (US$)Per month, billed monthly (US$)Limit
Tabitha.ai, 2023, founder's statement19free trial, then one plan
Free Starter, 2026, later owner0015 credits on signup
Tabitha Pro, 2026, later owner15.9919.99150 credits a month
School and district licence, 2026, later ownerprice on request, purchase orders

The founder said in July 2023 that the app was a subscription at US$19 a month with a free trial, and that a free-trial user cost her about US$0.25 to serve. The 2026 page, run by a later owner, sells a free plan with 15 credits and a Pro plan at US$19.99 a month with 150 credits, where one credit is one graded submission, one report-card comment or one AI-detection scan; it says yearly plans are 20% cheaper, which this page computes as US$15.99 a month. School and district licences are priced on request. Payments in 2026 go through Lemon Squeezy, not the Stripe setup the founder described.

02

The timeline

From a tool built for one teacher to a sale on a marketplace for tiny businesses, in about five months of part-time work. Each entry carries its grade.

  1. April to November 2022CBarker-DeStefano builds Shout About, a media-contacts app, on Bubble with a co-founder, pre-sells 20 subscriptions on Twitter and sells the business for US$15,000 after a broken leg stops her work. She describes a venture studio that builds small apps and sells them. [2]
  2. 27 January 2023CThe link-preview image of tabitha.ai is uploaded to Bubble, by the time stamp in its file name. The founder later says the first version, an AI detector for her partner's classroom, took about two weeks. [1, 3]
  3. 26 March 2023CFirst stored copy of tabitha.ai in the Wayback Machine: "The powerful AI teaching assistant that helps to improve educational experiences for everyone", a Bubble app. [3]
  4. Spring 2023DGenerative tools added: lesson plans, assessments, topic research, feedback drafts. Marketing is six weeks of scheduled social posts, YouTube videos and cold emails to teachers found on Apollo. About 40 to 50 sign-ups, 2 of them paying. [1]
  5. By 6 July 2023DListed on Tiny Acquisitions; interest within a day; sold for US$4,250 less a 9.5% fee to a buyer who wants to learn Bubble and has contacts in teaching. Revenue at the sale: US$38 a month. [1]
  6. 15 December 2023CThe site is still the same Bubble app with the same description, now under the buyer. [4]
  7. October 2026Ctabitha.ai sells a paper grader, a report-card writer and an AI detector at US$19.99 a month through Lemon Squeezy, with district licences. The page claims 12,000 teachers; nothing on it confirms the claim, and it belongs to a later owner. [5, 6]
03

Why it worked

Four causes, in the order they mattered. Each one is a reusable lesson.

01

Build for one known user, in two weeks

The first version solved one problem for one teacher, the founder's partner: is this essay written by an AI model? The founder said in July 2023 that she built it in about two weeks of part-time work with no announcement, because the only open question was whether she could make the AI integrations work in Bubble. Validation came after the build, through sign-ups.

02

No code, so the app itself is the asset

Bubble for the app, Stripe for payments, PriceWell to connect the two, MailerLite and Postmark for emails, FreshDesk for help articles, Reflio for affiliates, Apollo for teacher email addresses. A buyer who cannot program can run and extend such a stack, and the founder said the buyer wanted exactly that: a working Bubble app to learn from.

03

Sell at the start, not at the end

The founder said she likes finding problems and getting the first users, not the long growth stage. She calls what she sells a business in a box: a built app with early sign-ups that another entrepreneur takes over. Her stated US$4,250 for an app with US$38 a month of revenue is a price for the build and the idea, not for the income.

04

A marketplace for tiny deals

Tiny Acquisitions lists businesses too small for the larger marketplaces. The founder said the listing drew interest within a day because it was a no-code product using AI for a specific problem, and that the platform fee was 9.5%.

How it was built

These details come from the founder's July 2023 interview and from the archived page code. They are grade D unless the page code shows them. [1, 3]

  • Bubble, a visual no-code builder, for the whole application: sign-up, log-in, the detection tool, the generative tools, saved and printable reports. The archived page code carries Bubble's page-load markers.
  • AI detection and generative AI connected through outside APIs; the founder does not name the providers.
  • Stripe for payments, with PriceWell to manage the Stripe-to-Bubble link.
  • MailerLite for welcome and marketing emails, Postmark for transactional emails such as password resets, FreshDesk for support articles, Reflio for an affiliate program.
  • Apollo to find teachers' email addresses for a cold email series; social media accounts filled with six weeks of scheduled content; YouTube videos of use cases.
  • The founder said a free-trial user cost her about US$0.25 to serve, which set the cost of the free tier.

Tools named in the sources

Each one opens its page in the Tools tab, with every story that uses it.

04

The dead that came before

Five companies in the Loot Drop database, all from the same 2010 accelerator wave, sold teachers tools for grading, lesson planning and classroom records, and all closed by 2013. The values below are from that database and are not independently verified.

Died 2010 to 2013Burned US$500K

ClassroomIQ

The promise

Digital grading of paper exams, so teachers keep paper tests but lose the hand marking.

Why it died

Schools adopted digital tools slowly, better-connected rivals reached them first, and the company did not move to automatic grading in time.

Died 2010 to 2013Burned US$150K

Blendspace

The promise

Drag-and-drop digital lesson plans built from web content.

Why it died

Google Classroom and Microsoft Teams absorbed the job, and a free plan for teachers never turned into revenue.

Died 2010 to 2013Burned US$250K

Edusight

The promise

Classroom records and assessment tracking for K-12 teachers.

Why it died

Slow school procurement, then simpler rivals such as ClassDojo and Seesaw won the teachers.

Died 2010 to 2013Burned US$120K

BrightLoop

The promise

Track and share each student's progress in real time.

Why it died

Google's free classroom tools covered the same need before the company could raise more money.

Died 2010 to 2013Burned US$750K

Socrative

The promise

Live classroom quizzes on phones instead of clicker devices.

Why it died

Free offerings from large platforms and slow school adoption; the technology was sold off while the stand-alone business was not sustainable.

What changed

All five predecessors raised money, hired, and tried to sell to schools, whose purchasing is slow and whose budgets are small; then Google shipped the same features free. Tabitha.ai never faced that test. The founder said in July 2023 that she sold it with 2 paying teachers, before any school sale and before the free tools from OpenAI and Google reached the same jobs.

Two things were different in 2023. A single person could build the whole app in two weeks with no-code tools and a language-model API, so the cost of trying was a few weeks of evenings rather than a seed round. And a marketplace existed for selling such an app at a four-figure price: the founder said Tiny Acquisitions found a buyer within a day. The money came from selling the built product, not from the teachers, who paid US$38 a month in total by her account.

05

Can one person copy it now?

The method, yes; the money, barely.

The founder's method is fully described and needs no capital: one real user, a two-week no-code build, a few weeks of scheduled posts and cold email, then a listing on a marketplace for tiny businesses. The founder said the result was US$4,250 before a 9.5% fee, for roughly five months of part-time work, and the only evidence is her own interview. A reader who copies the method should expect a price for a working app and an idea, not for revenue.

Where it is still open: Any profession with repeated writing work that the large free assistants do not yet package for it: a specific exam board, a specific report format, a specific language of instruction.

  1. 1Pick one person you know with a recurring task and build the smallest tool that does it, on a no-code builder with a language-model API.
  2. 2Charge from the first day with a free trial, and know what a trial user costs you per month.
  3. 3Schedule several weeks of posts in one sitting and send a short cold email series to the exact profession.
  4. 4If you do not want to run the growth stage, list the app on a small-business marketplace while it still has sign-ups and a working stack a buyer can learn from.

What to watch

The risks are the mirror image of the causes above.

What is not verified

The headline figure, a sale for US$4,250, rests on one source: the founder's own interview in the No-Code Exits newsletter, published on 6 July 2023. The same interview is the only source for the US$38 a month of revenue, the 2 paying users, the US$19 price and the 9.5% marketplace fee. The buyer is not named, the Tiny Acquisitions listing is archived only as a script-rendered page with no readable text, and no payment-processor profile exists. A reader could look for the buyer's own account of the purchase, a Tiny Acquisitions record of the closed listing, or a later statement by the founder that repeats the figures.

Two users is not a business

By the founder's account the app had US$38 a month of revenue when sold. The price paid, if accurate, was for the built app and the early sign-ups. Nothing on this page shows that the product could have grown.

The later site makes claims this page cannot check

In October 2026 tabitha.ai says that more than 12,000 teachers use it. That is a later owner's statement with no evidence on the page, and it has no bearing on the 2023 sale.

Free assistants cover the same jobs

ChatGPT, Gemini and the free teacher tools from large companies write lesson plans and feedback at no charge, and AI detectors are known to produce false accusations against students. Every predecessor on this page died when a free platform covered its job.

Sources

The grade next to each source is the best grade that the source can support. "Read in full" means that the text of the page was downloaded and the numbers were checked in it.

  1. D No-Code Exits newsletter, 6 July 2023: How Milly built and sold an AI teaching assistent
    Read in full. The founder's interview. The only source for the sale price, the fee, the revenue, the user counts, the price, the build time and the tool stack.
  2. C No-Code Exits newsletter, 3 November 2022: How Milly went from validation to acquisition with No-Code
    Read in full. The founder's earlier interview about Shout About. Source for her background, her venture studio and her first exit.
  3. C Wayback Machine: tabitha.ai on 26 March 2023, the earliest stored copy
    Read the page code. The page is a Bubble app rendered by script, so only the title, the description and the Bubble markers are readable. The capture list came from the Wayback index for tabitha.ai.
  4. C Wayback Machine: tabitha.ai on 15 December 2023
    Read the page code. Same Bubble app and description after the sale.
  5. C Tabitha.ai pricing page, read 23 October 2026, run by a later owner
    Read in full. Source for the 2026 plans, the credits, the yearly discount and the Lemon Squeezy payments.
  6. C Tabitha.ai home page, read 23 October 2026, run by a later owner
    Read in full. Source for the 2026 features and the unverified user claim.
  7. C Birkwood Ventures, the founder's studio site, read 23 October 2026
    Read in full. A one-page site for a London company offering startup consulting; it does not mention Tabitha.ai.
  8. C Loot Drop database export, 1,749 failed startups
    Local export. AI-written summaries, not independently verified. Source for the failed predecessors.

Back to all stories