Learnify Genie

• 

• 

• 

• 

• 

• 

• 

  architecture diagram

Inspiration

We wanted to make learning and storytelling more visual and fun. Reading long passages of text can feel dull, so we thought — what if AI could turn any story or concept into a short, illustrated comic? That’s how Learnify Genie was born — a tool that turns text into engaging visuals and quick flashcards to help people learn faster and enjoy the process.

What it does

Learnify Genie takes any text, summarizes it into key points, turns those into a 4-panel comic strip, and also creates Q&A flashcards for quick learning. In just a few seconds, it transforms written content into something visual, memorable, and interactive.

How we built it

We built Learnify Genie using Flask for the backend and integrated AWS Bedrock with Claude 3 for text summarization and storytelling, and Nova Canvas for comic-style image generation. The app follows a three-stage pipeline — summarizing text, scripting the comic, and generating images — before returning the final comic panels and flashcards via a simple API.

Challenges we ran into

Parsing and cleaning JSON outputs from large language models Staying within token and prompt size limits for long texts Maintaining visual and narrative consistency across all four comic panels Handling large image generation responses efficiently

Accomplishments that we're proud of

Successfully automated the entire pipeline from text → comic → flashcards Achieved consistent comic art style using Nova Canvas prompts Built a fully working Flask API that connects multiple Bedrock models seamlessly Created a tool that blends creativity, education, and AI storytelling beautifully

What we learned

We learned how to orchestrate multiple AI models effectively, how prompt engineering impacts both text and image generation, and how to make LLM outputs more structured and predictable. We also gained hands-on experience integrating Bedrock APIs in a real-world Flask application.

What's next for Learnify Genie

We plan to add:

A frontend dashboard for users to upload text, documents and view comics interactively Support for multiple art styles (manga, watercolor, minimalist, etc.) A “Learn Mode” where users can review AI-generated flashcards for memory retention Integration with educational platforms to make reading and learning more engaging

Built With

• amazon-web-services
• bedrock
• claude
• flask
• nextjs
• python
• react