• Developed the twitter backend system with high concurrency and high availability by using Python based Django
• Leveraged Redis and Celery to implement asynchronous task queue to optimize newsfeeds service
• Leveraged Redis and Memcached to cache user profiles, tweets, newsfeeds, number of comments and likes
• Utilized denormalization method to store the number of likes and comments to avoid N+1 queries in MySQL
• Leveraged HBase, Thrift, HappyBase to store friendships and Amazon S3 to store uploaded images
• Utilized Django Ratelimit to prevent malicious requests and implemented django notifications to send notifications
• Deployed to Amazon EC2 with configured Nginx and Gunicorn
