-
-
How web application look like
-
Entries in Firestore DB after entered in Frontend
-
High level Architecture Diagram
Inspiration
Diwali is a festival of joy and connection, and I wanted to use AI to help people send personalized greetings that truly reflect their feelings and traditions. The project combines serverless technology and generative AI to make this process creative, scalable, and accessible to everyone.
What it does
Diwali Dev Vibes lets users enter their preferences and generates unique, AI-powered Diwali greetings. It stores user data and greetings in Firestore, and delivers responses instantly via a web interface deployed on Cloud Run.
How we built it
I used Google AI Studio to design the greeting generation logic, integrated it with a Flask app, and connected Firestore for data storage. The app was containerized with Docker and deployed to Cloud Run for scalability and ease of access. Environment variables and Secret Manager were used for secure key management.
Challenges we ran into
Managing API keys securely, handling free tier quota limits for the Gemini model, and ensuring all dependencies worked smoothly in both local and cloud environments were some of the main challenges. Debugging SSL certificate issues on macOS also took extra effort.
Accomplishments that we're proud of
Successfully integrating generative AI with a server less architecture, deploying a fully functional app to Cloud Run, and seeing real user data and greetings stored in Firestore. The project demonstrates a modern, scalable approach to combining AI and cloud technologies.
What we learned
I learned how to securely manage secrets in cloud deployments, troubleshoot cloud and Python environment issues, and leverage Google’s AI models for creative applications. The experience deepened my understanding of server less architecture and cloud-native development.
What's next for Diwali & Dev Vibes
I plan to add more customization options for greetings, support other festivals, and improve the user interface. Scaling the app for more users and integrating additional Google Cloud services like Pub/Sub or BigQuery are also on the roadmap.
Log in or sign up for Devpost to join the conversation.