🦄 Unicorn.Rentals - AWS Bedrock Hackathon 2025

AWS Cloud Istanbul Bedrock Hackathon 2025 Submission

A magical AI-powered healthcare system for mythical creatures, built with Amazon Bedrock and Streamlit. This project demonstrates advanced AI capabilities including multimodal analysis, knowledge base integration, and creative content generation.

🌟 Project Overview

Unicorn.Rentals is a comprehensive healthcare platform for mythical creatures that leverages Amazon Bedrock's powerful AI models to provide:

• Health Complaint Analysis: AI-powered diagnosis of unicorn health issues
• Medical Knowledge Base: RAG-powered medical information system
• Multimodal AI Oracle: Text and image generation capabilities

🏗️ Architecture

┌─────────────────┐    ┌──────────────────┐    ┌─────────────────┐
│   Streamlit     │    │  Amazon Bedrock  │    │   Amazon S3     │
│   Frontend      │◄──►│                  │◄──►│ Knowledge Base  │
│                 │    │  • Nova Lite     │    │                 │
│  • UniSum       │    │  • Nova Pro      │    │ Medical Files   │
│  • WisdomCare   │    │  • Nova Canvas   │    │ • Anatomy       │
│  • Mystic Oracle│    │  • Titan Models  │    │ • Ailments      │
└─────────────────┘    │  • Claude 3      │    │ • Protocols     │
                       └──────────────────┘    └─────────────────┘

🚀 Features

1. 🩺 UniSum - Health Complaint Analyzer

• PDF Processing: Extract text from medical transcripts
• AI Analysis: Use Nova Lite for symptom analysis and severity assessment
• Report Generation: Generate comprehensive veterinary reports with Nova Pro
• Triage System: Automatic urgency classification (Low/Medium/High/Emergency)

2. 🧠 WisdomCare - Medical Knowledge Base

• RAG Implementation: Retrieval-Augmented Generation with Bedrock
• S3 Integration: Automated medical knowledge upload
• Multi-Model Support: Nova Lite, Nova Pro, and Claude 3 Haiku
• Interactive Queries: Natural language medical consultations

3. 🎨 Mystic Oracle - AI Content Generation

• Text Generation: Multiple model support (Nova, Titan, Claude, Llama)
• Image Creation: Nova Canvas, Stable Diffusion, Titan Image Generator
• Model Access Checker: Real-time model availability verification
• Multimodal Capabilities: Combined text and image generation

🛠️ Technology Stack

• Frontend: Streamlit
• AI/ML: Amazon Bedrock
  • Amazon Nova Lite (multimodal analysis)
  • Amazon Nova Pro (advanced text generation)
  • Amazon Nova Canvas (image generation)
  • Amazon Titan Text & Image models
  • Anthropic Claude 3 models
  • Meta Llama models
• Storage: Amazon S3
• Knowledge Base: Amazon Bedrock Knowledge Bases with Titan Embeddings V2
• Language: Python 3.8+
• Libraries: boto3, streamlit, PyPDF2, Pillow

📋 Prerequisites

1. AWS Account with appropriate permissions
2. Amazon Bedrock Model Access:
   • Amazon Nova Lite (amazon.nova-lite-v1:0)
   • Amazon Nova Pro (amazon.nova-pro-v1:0)
   • Amazon Nova Canvas (amazon.nova-canvas-v1:0)
   • Amazon Titan Text Embeddings V2
   • Other models as needed
3. Python 3.8+
4. AWS CLI configured with appropriate credentials

🚀 Quick Start

1. Clone the Repository

git clone https://github.com/ilkan/aws-cloud-istanbul-bedrock-hackathon-2025.git
cd aws-cloud-istanbul-bedrock-hackathon-2025

2. Install Dependencies

pip install streamlit boto3 PyPDF2 Pillow pandas

3. Configure AWS Credentials

aws configure
# Enter your AWS Access Key ID, Secret Access Key, and region (us-east-1)

4. Enable Bedrock Models

1. Go to AWS Bedrock Console
2. Navigate to Model access under Bedrock configurations
3. Click Modify model access
4. Enable required models:
   • ✅ Amazon Nova Lite
   • ✅ Amazon Nova Pro
   • ✅ Amazon Nova Canvas
   • ✅ Amazon Titan Text Embeddings V2
   • ✅ Other models as needed
5. Submit the request

5. Run the Applications

UniSum - Health Analyzer

cd environment/GameDay/ml_repo
streamlit run 1_UniSum.py

WisdomCare - Knowledge Base

cd environment/GameDay/ml_repo
streamlit run pages/2_WisdomCare.py

Mystic Oracle - AI Content Generator

cd environment/GameDay/mystic_code/mystic_code
streamlit run bedrock_app_st.py

📖 Usage Guide

UniSum Health Analyzer

1. Upload the provided transcript1.txt as a PDF
2. Click Analyze Health Complaint to get AI-powered diagnosis
3. Generate a comprehensive veterinary report
4. Review triage recommendations

WisdomCare Knowledge Base

1. Create an S3 bucket for medical knowledge
2. Upload sample medical files
3. Create a Bedrock Knowledge Base in the AWS Console
4. Enter the Knowledge Base ID in the app
5. Ask medical questions and get AI-powered answers

Mystic Oracle

1. Choose between text generation, image creation, or both
2. Select your preferred AI model
3. Enter prompts and generate magical content
4. Download generated images

🎯 Hackathon Highlights

Innovation Points

• Multimodal AI: Combines text, image, and document processing
• RAG Implementation: Advanced knowledge retrieval system
• Model Diversity: Utilizes multiple Bedrock models for different tasks
• Real-world Application: Practical healthcare use case with whimsical twist
• User Experience: Intuitive Streamlit interface with magical theming

Technical Achievements

• PDF Processing: Automated transcript extraction and analysis
• Dynamic Model Selection: Runtime model availability checking
• Error Handling: Comprehensive error management and user guidance
• Scalable Architecture: Modular design for easy extension

📁 Project Structure

aws-cloud-istanbul-bedrock-hackathon-2025/
├── README.md
├── environment/
│   └── GameDay/
│       ├── ml_repo/
│       │   ├── 1_UniSum.py              # Health complaint analyzer
│       │   ├── pages/
│       │   │   └── 2_WisdomCare.py      # Medical knowledge base
│       │   └── transcript1.txt          # Sample medical transcript
│       └── mystic_code/
│           └── mystic_code/
│               └── bedrock_app_st.py    # AI content generator
├── data/                                # Sample data files
├── logs/                               # Application logs
└── .kiro/                              # IDE configuration

🔧 Configuration

Environment Variables

export AWS_DEFAULT_REGION=us-east-1
export AWS_PROFILE=your-profile-name  # Optional

Model Configuration

The applications automatically detect available models and provide guidance for enabling access to required models.

🎨 Sample Use Cases

1. Emergency Triage: Analyze urgent unicorn health complaints
2. Medical Research: Query comprehensive medical knowledge base
3. Content Creation: Generate magical artwork and stories
4. Educational Tool: Learn about AI model capabilities

🏆 Awards & Recognition

This project was developed for the AWS Cloud Istanbul Bedrock Hackathon 2025, showcasing:

• Advanced Amazon Bedrock integration
• Creative AI applications
• Practical problem-solving with whimsical presentation
• Multi-model AI orchestration

🤝 Contributing

This is a hackathon project, but contributions and improvements are welcome!

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Submit a pull request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• AWS Bedrock Team for providing powerful AI models
• AWS Cloud Istanbul for hosting the hackathon
• Streamlit for the excellent web framework
• The magical unicorn community for inspiration 🦄

📞 Contact

Team: Unicorn.Rentals Development Team
Hackathon: AWS Cloud Istanbul Bedrock Hackathon 2025
Repository: GitHub

---

Built with ❤️ and a touch of magic during the AWS Bedrock Hackathon 2025 ✨