Every author knows the feeling, finishing a book is only the beginning, the real challenge is ensuring it finds readers.
For decades, book discovery has relied on publishers, booksellers, reviewers and, increasingly, social media influencers. Yet despite millions of titles being available online, readers often complain that finding their next favourite book has become harder rather than easier.
Michelle Zaffino believes the system is broken. A former magazine journalist turned digital librarian, software developer and founder, she has spent the past seven years building MyLibrarian, an AI-powered recommendation platform that takes an unusual approach to solving the problem.
Rather than relying primarily on purchasing habits or social media trends, the platform uses one of publishing’s oldest and most trusted resources – the librarian.
“I write code and words,” she says, reflecting a career that has crossed journalism, information science, software engineering and publishing. “It’s my mission to help authors have their books discovered, so we can sell more books and have more time to write.”
After working as a journalist in New York for publications including Marie Claire, Redbook, Elle, Esquire and Sports Afield, Zaffino relocated to San Francisco, where she completed a Master’s in Information Science before teaching herself software development and machine learning.
The catalyst was surprisingly simple. She wanted to research family history for a historical novel she was writing. The deeper she immersed herself in libraries and information science, however, the more she became convinced that the biggest challenge facing authors wasn’t writing books—it was helping readers find them.
“I decided that book discovery is broken,” she explains and that observation eventually became her startup.
Trust over trends
Book recommendation engines are hardly new. Amazon has long relied on purchasing behaviour. Goodreads encourages community reviews. TikTok’s BookTok phenomenon has transformed obscure novels into international bestsellers.
Zaffino believes another group deserves far greater influence.
“Librarians have just as much clout as book clubs,” she says. “They’re not going to steer you wrong.”
The idea behind MyLibrarian is straightforward. Instead of recommending books because they are trending or heavily marketed, the platform aggregates librarian reviews and metadata into what Zaffino calls the Librarian Brain, a recommendation engine built from professional curation rather than popularity alone.
Users can then receive recommendations and be directed towards borrowing or purchasing titles through integrated partners. For authors, particularly independent authors, the concept is especially appealing.
Traditional publishing remains highly competitive, while self-publishing offers unprecedented access but little guarantee of visibility. Even authors with established audiences frequently struggle to cut through an increasingly crowded marketplace.
Zaffino understands both worlds. Despite having publishing contacts from her years in New York, she eventually chose to establish her own independent publishing company and publish through platforms such as Draft2Digital and Amazon KDP.
“It was all about your follower count and your platform,” she recalls of the traditional route. “Even if you’re traditionally published, you still need to build your platform.”
Building an AI startup the hard way
Like many founders, Zaffino’s journey has been less glamorous than headlines often suggest.
Over seven years she has raised around $400,000, enough to build an early version of the platform while operating with an exceptionally lean international team of contractors spanning Germany, Kosovo and Sri Lanka. She remains the company’s only full-time employee.
The MyLibrarian app is currently available through TestFlight while further funding is sought to accelerate development.
The emergence of generative AI has undoubtedly changed the landscape. But, Zaffino is keen to distinguish between today’s public fascination with AI and the technologies that have quietly underpinned information retrieval for years.
“We’ve been building with AI for a decade,” she says. “People have always been using AI; they just don’t even know it.”
She describes herself as an “AI for Good” advocate, believing that combining artificial intelligence with librarians’ expertise offers a more trustworthy alternative to purely engagement-driven recommendation systems.
More than another book app
Perhaps the most intriguing aspect of MyLibrarian is that it isn’t simply trying to build another reading app, it is attempting to build a community.
Alongside recommendations, the platform is developing book clubs, librarian ambassadors, creator content, community features and eventually peer-to-peer payments for writers and creators. More than 11,000 people have already registered their interest ahead of a wider launch.
Whether that community becomes the missing link between readers, librarians and authors remains to be seen, but in an era increasingly dominated by algorithms optimised for clicks and attention, Zaffino is making a different bet.
It’s Zaffino’s argument that the future of AI-powered discovery may depend less on replacing trusted human judgement than amplifying it. For readers searching for their next favourite book, and for authors hoping to be found, that is a very powerful premise.
Michelle Zaffino
Founder, MyLibrarian / In the Stacks
www.IntheStacks.tv
www.michellezaffino.com
Check out my novels:
http://amazon.michellezaffino.com

