City officials in Belagavi uncovered a troubling insight: only 4% of students read actively, while 96% did not. Investigations revealed a surprisingly common cause—students found books boring, irrelevant, or too complex to understand.

Instead of launching conventional awareness campaigns, the city decided to rethink the book itself. What if the book could learn about the reader—and evolve to suit them?

How the AI-Enhanced eBook Works

The core innovation lies in using deep learning and analytics embedded directly inside EPUB-format eBooks. Here’s what makes them unique:

• Behavioral Analytics: The eBook tracks user engagement, mood, and interest in real time.
• Dynamic Story Adaptation: Based on user input, story elements and paragraphs are reshuffled to suit the reader’s attention and enthusiasm.
• Personalized Vocabulary: The book automatically simplifies or enriches vocabulary depending on how often the reader looks up words.
• Integrated Micro-Assessment: For exam-prep editions, subtle questions are added to gauge proficiency and adjust difficulty without disrupting flow.

All this is done using AI algorithms that sit inside the eBook—no constant internet required.

Learning in Five Languages, Even Offline

The platform supports five languages through smart eFlashcards, which feed data into the deep learning algorithm. This ensures that students who are translating from their mother tongue to learn new content aren’t penalized by linguistic barriers.

The AI helps the system avoid literal or poor translations, creating a more natural learning experience. And since these tools work offline, they’re fully usable in low-connectivity rural or semi-urban zones.

Early Gains: Better Readers, Sharper Thinkers

An initial pilot with 70 students prepping for competitive exams showed impressive results:

• 12% improvement in reading speed in just two weeks
• 18% improvement in test scores
• Increased confidence and comprehension, particularly in learners under age 30

Even teachers benefited: a sample group of 20 educators reported better pedagogy and learning retention after being trained on the platform.

These eBooks are now being deployed beyond the pilot, with requests from schools across Karnataka to install in-campus replicas of the smart e-library.

Scaling Smart: From One Library to Statewide Adoption

Currently, Belagavi’s e-library operates across three floors and caters to pre-primary and primary students. As of now:

• Every AI-generated book includes the reader’s name and image, making them the protagonist.
• All eBooks are free for citizens, including special editions created in the name of the State Chief Secretary.

There is growing interest in expanding this approach to school libraries and academic test prep centers. Schools are even willing to pay a subscription fee for access—testament to the perceived value.

City-Scale AI, Without the Data Bloat

One major learning from the deployment: AI doesn’t need to be cloud-reliant. Offline algorithms using lightweight data collection have been crucial for ensuring:

• The device memory isn’t overloaded
• Users maintain privacy
• Algorithms remain functional without active connectivity

These lessons could guide similar deployments across other smart cities under Viksit Bharat @2047.

A Smarter Way to Read—and Learn

Belagavi’s Active AI eBooks prove that even something as traditional as reading can be revolutionized with the right blend of behavioral data and cognitive science. By making books that read their readers, this initiative is unlocking new ways to personalize learning, boost exam readiness, and revive interest in reading across all age groups.