Inspiration
Have you ever struggled to finish a book—or wished your documents could speak to you?
I love reading, but with each passing day, my daily routine and the rat race pull me further away from my bookshelf. By the end of the day, I often find myself binging a series rather than flipping pages. That’s when the idea struck: What if I could transform the books in my personal library into cinematic story experiences or interactive tutorials?
As a software engineer, I'm constantly learning—reading documentation, exploring new tools, and watching tutorials. But with limited time and attention, I can’t do it all.
I drew inspiration from the immersive magic of Harry Potter, and imagined a future where books didn’t just talk to you—they guided you, taught you, and transported you.
What it does
Litinkai transforms static book content into dynamic AI-generated learning and entertainment experiences.
You can upload content in .pdf, .epub, or .docx format—or even type in your own stories—and Litink will convert it into:
🎓 Interactive Learning Mode: AI-curated lessons, smart quizzes, and verifiable blockchain credentials.
🎬 Entertainment Mode: Choose-your-path cinematic stories with AI-generated voices, scenes, and collectibles.
Whether you're a curious learner or a creative storyteller, Litink makes every page come alive.
How we built it
Litink was built with a combination of prompt-driven development (using Bolt.new and Cursor IDE) and a powerful modern tech stack:
⚙️ Backend: FastAPI
🖼️ Frontend: React
📈 Database: Supabase
💳 Payments: Stripe
🧠 AI & LLMs: OpenAI, ElevenLabs, Tavus, Bolt.New, Cursor
🎞️ Generative Video: Kling.ai
🌐 Deployment & Domain: Netlify + IONOS
We used OpenAI for text-to-lesson and script generation, Tavus for voice/video narration, ElevenLabs for voicenarration and Kling.ai for cinematic video content. Blockchain credentials were powered by Algorand.
Challenges we ran into
Turning static books into truly dynamic, interactive experiences
Building a flexible RAG (Retrieval-Augmented Generation) system on a tight deadline
Integrating multiple AI video tools with different limitations
Short-form video generation using Tavus was helpful for conversational learning, but less suited for long, structured tutorials like those on Udemy or Coursera
Kling.ai produced great cinematic content, but we quickly exhausted our usage credits
Accomplishments that we're proud of
✅ The MVP works!
🎯 Proved that books can be transformed into personalized, interactive content
💻 Built my first end-to-end fullstack AI application
🔗 Successfully integrated LLMs, speech, and blockchain into one cohesive product.
What we learned
With AI, imagination is the only limit.
With Bolt.new, even solo builders can ship production-ready tools.
Merging learning and storytelling creates powerful, engaging content.
What's next for https://www.litinkai.com/
📈 Launch to early users
💸 Apply for grants and fundraising
⚙️ Improve video generation quality and tutorial structure
🌍 Expand multilingual and accessibility features
🧠 Integrate deeper interactivity and character-driven narratives

Log in or sign up for Devpost to join the conversation.