Achamin

An innovative web application that transforms how people discover, understand, and emotionally connect with global cultures

Comment

Inspiration

Our team was inspired by the growing disconnect between people and diverse cultural traditions in our increasingly digital world. We noticed that while technology has made the world smaller in many ways, it has also created barriers to deep cultural understanding. Museums and cultural institutions struggle to create immersive experiences that truly engage visitors emotionally with artifacts and traditions.

We asked ourselves: "What if AI could help bridge this gap by creating personalized, emotionally resonant cultural experiences?" This question led us to develop Achamin, a platform that transforms how people discover and connect with global cultures through the power of generative AI.

What it does

Achamin is an innovative web application that transforms how people discover, understand, and emotionally connect with global cultures. When users upload an image of a cultural artifact, Achamin:

  1. Analyzes the image using Amazon Rekognition to identify cultural elements
  2. Generates rich cultural context using AWS Bedrock and Claude to create a narrative about the artifact
  3. Creates an audio narration using Amazon Polly with 8 different narration styles and voices
  4. Adds background music that matches the cultural context and mood
  5. Delivers an immersive, multi-sensory experience that helps users form an emotional connection with the cultural artifact

The application supports 8 different narration styles (storytelling, educational, conversational, poetic, historical, personal, analytical, inspirational) and uses neural voice technology for superior audio quality.

How we built it

We built Achamin using a serverless architecture on AWS:

  1. Frontend: HTML/CSS/JavaScript for a responsive, intuitive user interface
  2. Backend: AWS Lambda functions written in Python for serverless processing
  3. AI Services:
    • Amazon Rekognition for image analysis
    • AWS Bedrock with Claude for cultural context generation
    • Amazon Polly for text-to-speech narration
  4. Storage:
    • Amazon S3 for storing images, audio files, and background music
    • DynamoDB for metadata storage
  5. API: Amazon API Gateway for RESTful API endpoints
  6. Orchestration: AWS Step Functions for complex workflows

We implemented a modular architecture with separate components for image analysis, cultural context generation, audio production, and content delivery. This approach allowed us to iterate quickly and enhance each component independently.

Challenges we ran into

  1. CORS Configuration: We encountered significant challenges with Cross-Origin Resource Sharing (CORS) when integrating our frontend with the API Gateway and S3 buckets. We had to implement multiple layers of CORS headers and create special error handling to ensure proper cross-origin requests.

  2. AI Prompt Engineering: Creating prompts that generated culturally accurate and sensitive narratives required extensive experimentation and refinement. We needed to balance creativity with factual accuracy.

  3. Audio Synchronization: Coordinating the narration with background music while maintaining high audio quality was technically challenging. We had to develop custom audio processing logic.

Accomplishments that we're proud of

  1. Cultural Sensitivity: We developed an AI system that generates culturally respectful and informative content, avoiding stereotypes and oversimplifications.

  2. Technical Integration: Seamlessly integrated multiple AWS AI services (Rekognition, Bedrock, Polly) into a cohesive application.

  3. Scalable Architecture: Built a serverless architecture that can handle varying loads efficiently and scale automatically.

What we learned

  1. AI Prompt Engineering: We gained expertise in crafting effective prompts for generative AI models to produce specific types of content.

  2. Serverless Architecture: We deepened our understanding of building and deploying serverless applications on AWS.

  3. Cultural Computing: We learned about the challenges and opportunities in using AI to represent and interpret cultural artifacts.

  4. Cross-service Integration: We developed skills in integrating multiple AWS services into a cohesive application.

  5. Error Handling in Distributed Systems: We learned techniques for robust error handling across distributed serverless components.

What's next for Achamin

  1. Cultural Expert Collaboration: Partner with cultural institutions and experts to enhance the accuracy and depth of our cultural narratives.

  2. AR/VR Integration: Add augmented and virtual reality features to create even more immersive cultural experiences.

  3. Multilingual Support: Expand to support narratives and audio in multiple languages.

  4. User Communities: Build community features that allow users to share and discuss cultural discoveries.

  5. Educational Partnerships: Collaborate with educational institutions to create curriculum-aligned cultural learning experiences.

Share this project:

Updates