India’s homegrown regional language social media platform ShareChat has built India's largest content ecosystems on Google Cloud.
The ecosystem brings together more than 180 million monthly active users who share and view videos, images, GIFs, microblog posts, and songs in over 15 different Indian languages.
ShareChat’s short video platform, Moj also has over 160 million monthly active users with the largest music libraries available for content creators in India.
ShareChat and Moj’s senior vice president of engineering, Gaurav Bhatia, told the Google Next 2022 event that the company has built its technology stack on a cloud-first strategy with Google’s advanced tools and integrated offerings to massively scale the services.
“We could grow 10 times more in scale without impacting developer productivity and feature velocity,” he added.
The firm migrated its infrastructure from Amazon Web Services (AWS) to Google Cloud to cut expenses on backend technology in 2020.
The company required tools to accelerate and improve the app development cycle.
“We looked for a cloud infrastructure with agile performance, cost, better cloud efficiency and auto scalability,” he added.
Google solutions
Sharechat has deployed a combination of managed services like Google Kubernetes Engine (GKE), Pub/Sub for data pipelines, Spanner for real-time app serving workloads, BigTable for less-indexed databases, Dataflow and BigQuery for analytics.
It has also used Cloud CDN to distribute high-quality content at low latency to consumers.
Bhatia said the firm has built its microservices architecture on a Kubernetes (k8) cluster hosted in the Google Cloud to break down each functionality into smaller components and scale "independently".
This helped the firm roll out new or updated services to a smaller percentage of consumers initially as a pilot to avoid errors.
The company also relies on a platform-as-a-service model to have central functions, providing enhanced infrastructure for teams building product functionality.
The firm has an engineering team to build abstractions on Google services like Database, Queue and Cache Drivers.
While services are built and deployed in a CI/CD model, their teams have also built internal tools like Atlas, that allow developers to max out the benefits of Google Cloud in a "few clicks". This included a machine learning infrastructure on the GCP for personalised recommendations.
Sharechat has shifted from open source server environment Node.js to GoLang that had resulted in "good" cost savings on infrastructure. It has also helped them easily monitor and maintain scalability.
On the future plans, Bhatia said, the company will work to rewrite the entire product stack using GoLang and explore other Google Cloud services in security and disaster recovery areas.
