Welcome to Books Butterfly AI

Books Butterfly AI is an AI company building a Real World Model

Our Books Butterfly Real World AI Model is built on reader behavior data from 2017 to 2026 (gathered from 4.3 million readers) and book marketing and book curation experience from 2008 to 2026.

***

We are building a Books & Publishing focused Real World Model that focuses primarily on ->

A) Writing The Right Book for The Right Market. This is called PMF AI (Product Market Fit AI)

B) AI powered Publishing with Zero Human Participation and Minimal Human Oversight. This is called ZeroPubAI

C) Removing the entire 'Book Store' construct and replacing it with an Intelligent AI Assistant which would understand a reader's taste in books even better than the reader. We call our AI reading assistant Lily AI

*****

A) Writing the Right Book for the Right Market - Details on PMF AI (Product Market Fit AI)

Writing The Right Book for The Right Market is a NP Complete problem and we are not going to pretend we can successfully solve it. At the same time, we can help Publishers and Authors get as close as humanly or AI-ly possible to finding Product Market Fit.

The Books Butterfly AI Real World Model will help Publishers and Authors with

1) Deciding what Market to Write & Publish Books For 

2) Determining the Level of Competition in the Market they are going into

3) Figuring out The Right Product for this 'Right Market'

4) Trends in Publishing and whether or not they affect this 'Right Market' they have chosen

5) The stability and permanence of this 'Right Market'

Additionally, and equally importantly, the Books Butterfly AI Real World Model will help Publishers and Authors with marketing and visibility and sales and customer retention

1) Perhaps Most Importantly -> Making the jump from a Book that is Good to a Book that is Very Good and will Sell in the Market

Please Note: The Jump from Very Good to Great is completely based on the writing skills of the author and how much they are willing to improve their understanding of the market and their writing level. We can only help you grow from Good to Very Good. The jump to Great is your own personal journey.

2) Budgeting your time and money and resources. Perhaps the biggest mistake new Publishers and new Authors make is that they spend 80% of their budget at the very start. At the very start is when they have neither the experience nor the skills to use their budget wisely

3) Budgeting your time and money and resources based on the quality level of the book

While AI in general and Books Butterfly in AI cannot write a great book, it can certainly assess and determine when a book is far off from being 'great' and far off from being 'marketable'.

We are going with an AI Real World Model for Publishing and Books because unlike areas such as legal research and programming, writing is a deceptively hard mix of being an artist and being a scientist and being a market researcher. Existing LLM models cannot capture this multi dimensional world of books with brute force machine learning.*****

*****

B) AI powered Publishing with Zero Human Participation and Minimal Human Oversight - ZeroPubAI

As opposed to many AI startups that want to remove the 'human' part of creation, we want to remove the 'Publishing' part of Books Publishing and replace it with AI Publishing. Humans would still write the books.

While AI is not yet (and perhaps never) able to write with the originality and human spirit of human authors, AI can do the vast majority of the rest of Publishing very efficiently.

A book in its writing, formatting, book cover, description, choice of book categories reveals its level of quality

After running over 30,000 book promotion packages successfully (for a combined $3.5 million worth) we've found the same indicators of whether or not an author will succeed, and whether or not a book will succeed.

Most (and hopefully all) of these can be translated into AI tasks.

The final step is the Books Butterfly AI Real World Model which would be able to tell which books are the most likely to succeed. It would elevate 'Publishing of Books' and decision making of which books to publish, from an art to both a science and an art.

Perhaps equally importantly, it would give Authors and Publishers invaluable feedback (as a service, not free) on what they can do to increase their books' chances of finding a market and a solid readership.

*****

C) A completely powered by AI Book Store i.e. An AI Agent that knows the readers' tastes even better than the reader does -> Lily AI

AI can replace the eBook store and the Book store completely.

By their very nature Book store and eBook stores are trying to be 'everything to everyone' and are the exact opposite of a great customized experience. This is especially true for Books where readers have their favorite 1 to 4 book genres and never touch any books outside those genres.

While people have their preferences in movies and dining and music and games, they still dabble in many different categories.

In stark contrast, the vast majority of readers have their preferred 1 or 2 or 3 or 4 book genres (usually just 1 or 2) and they never ever venture outside those genres.

3 very good examples are - Steamy Romance readers, True Crime readers, Biography & Memoir readers. Let's consider True Crime readers. They will walk into a book store, or visit an online book store, and never look at any book category other than True Crime. The rest of the book store is just a distraction.

A Real World Model Reading AI Assistant would be able to create a customized book store, and in reality a customized world for the reader where they see nothing except what they love to read.

Perhaps equally importantly, it could do all the hard work of finding which True Crime events have been transformed into quality True Crime books. Thereby, the very hard 'searching for a good True Crime book' aspect would be handled by the Reading AI itself

One amazing aspect of Lily AI is that existing bookstores and ebook stores are trapped in their current models. They introduce various inefficiencies in book discovery so that they can sell expensive, ineffective, and hard to understand marketing packages. In stark contrast, Lily AI focuses on removing all friction and resistance between a reader and their next book.

*****

FAQ on Books Butterfly AI

1) If you have experience in Real World Models and would like to work at a company that is 100% focused on building the best Real World Models with zero politics and zero drama please get in touch at ai@booksbutterfly.com

2) We do not build any products for Governments or Militaries

3) We are a very small and lean company and please don't mind if we can't get back to you. If we feel there is a very good fit we will definitely get in touch. If we don't feel there is a good fit, we might not have time to respond

4) We are not looking for funding. We are happy to hear your suggestions and answer questions. However, please don't get upset regarding funding as we are not looking for funding. You can reach us at aiq@booksbutterfly.com

5) We do not work with or in the area of replacing human authors with AI authors. Sorry, we will not partner with you if you intend to replace human authors with AI. Not only do we think this is impossible, there is also copious amounts of evidence that readers will not buy or read AI written books. While someone at some point might build an AI engine that writes as well as humans, human readers will still reject it because a massive part of the reading experience is connecting with a fellow human (the author).

6) If you have a large amount of readers, or a lot of data on reader purchases and behavior, we are very interested if you are open to an exclusive deal. At the moment we only do exclusive deals which will give us access to readers or reader behavior data that no one except us can leverage

7) We do not buy Data Sets. At the moment we are not looking for data labelers or data companies. We do appreciate your interest. We just happened to get lucky that in the process of our curation and marketing efforts we already were labeling the rather large amount of data on books and publishing we were gathering - All the way since 2012/2013/2014.

*****

Books Butterfly AI's team includes people with experience from FAANG companies, and Enterprise SaaS companies with over $5 billion exits.

The founders have Computer Science degrees from world class institutions and fully understand the underlying technology with multiple decades of experience in Neural Networks, Evolutionary Algorithms, and Data Science.

*****

Our Experience and Data for Publishing & Books includes

1) Book Curation

Curation of Books since Dec 2008 to May 2026 (and ongoing). At the moment we curate and market and promote over 200 books and paperbacks for our 4.3 million readers every single day (1.15 million monthly active and 675,000 daily active).

2) Book Promotion & Marketing

We've successfully run $3.5 Million of book promotion packages. We have a miniscule 0.35% dispute rate (36 disputes in our last 10,000 book promotion packages).

We've covered both Fiction and Non Fiction with our strongest book categories being

Fiction -> Romance, Steamy Romance, Mystery, Thriller, Historical Fiction, Fantasy, LitRPG, Suspense, Adventure, Women's Fiction.

Non Fiction -> Business, Biographies & Memoirs, True Crime, History, Technology, Cookbooks, Health & Fitness, New Age, How To, DIY.

3) Data on Book Trends

We have Data on Publishing & Book Trends from June 2015 to Apr 2026 (and ongoing). Apart from the eBook stores and Bookbub.com, there is no other company that has the amount of data and insight from real customers that we have gathered from our 4.3 million readers.

We market and promote and curate (40% to 60% ratio approximately) over 200 ebooks and paperbacks every single day across Amazon Kindle ebook store, Amazon.com for paperbacks, B&N Nook and Apple iBooks and Kobo for eBooks. We have daily data from each of these ebooks and paperbacks and how they performed.