Go Back

AWS Cloud-Based Movie Recommendation System

Cloud-native movie recommendation app with Go microservices, React, AWS infrastructure, SageMaker recommendations, and LLM-powered suggestions.

This project is a cloud-native movie recommendation system with a web frontend, authenticated user flows, Go microservices, ML-powered recommendations, and natural-language movie search. The deployed site is available at findamovie.online.

Users can sign up, browse and search movies, like titles, and receive recommendations from a KNN model served through SageMaker. The app also supports AI-powered prompts through Gemini for requests such as finding movies with a similar story, mood, or theme.

Architecture:

  • Frontend: React web app deployed on AWS Amplify.
  • Auth: Amazon Cognito issues ID tokens for protected API requests.
  • API: API Gateway routes authenticated traffic through an Application Load Balancer.
  • Backend: Go movies-service and users-service deployed on Amazon ECS with Fargate and EC2 comparison work.
  • Data and ML: DynamoDB stores user preferences, TMDB powers catalog data, S3 stores model artifacts, and SageMaker serves recommendation inference.
  • Ops: ECR stores service images and CloudWatch supports logs and monitoring.