Bobble.AI’s AWS to GCP Migration Resulted in Enhanced Scalability, Performance, and Reduced Cost
Niveus Facilitated the End-to-end Migration Process with Strategic Cloud Solutions
Business Impact
150 milliseconds
or less latency achieved
99.95% success rate
for the APIs
87% reduction
in infrastructure provision & scaling time
15% reduction
in overall cost
The Client
Bobble.AI, a leading technology innovator, has developed the Bobble AI Keyboard, a user-centric mobile application designed to revolutionize communication. Headquartered in Gurugram, India, Bobble.AI is a dynamic company with over 300 dedicated employees. With branches across three continents, including offices in Southeast Asia and a growing presence in Europe, Bobble.AI is rapidly expanding its global footprint. The company supports over 150 languages and is equipped with advanced autocorrect and prediction technology.
Business Context:
With chat becoming the primary communication channel, Bobble.AI offers a unique solution for brands to achieve prominent placement through branded stickers and content integrated within the smartphone keyboard during conversations. This organic sharing by users signifies their genuine interest and preference for the brand. Bobble.AI was aiming to leverage a robust cloud platform offering advanced features and scalability to deliver this innovative marketing approach more efficiently and effectively.
Challenge
Bobble.AI faced challenges with their current cloud environment, including high costs and limitations on scaling resources. To address these issues, they decided to migrate to a new cloud provider for cost optimization and to deliver a better customer experience by reducing latency.
Opportunity:
The primary goal was to gain access to an optimized cloud infrastructure. Considering this goal, Niveus suggested migrating the client’s backend applications and services to Google Cloud, which offers a smooth transition without disruption to business and has minimal latency. Niveus’ team has the necessary skills to leverage the full potential of GCP’s storage and database capabilities. This AWS to GCP migration aimed to reduce overall costs while ensuring a secure process with minimal disruption, ideally achieving zero downtime for all data and services. Additionally, Bobble.AI expressed interest in further utilizing Niveus’ cloud operations services.
Business Solution
Niveus successfully migrated Bobble AI Keyboard’s applications from AWS to GCP. The team successfully set up the staging and production environments for Bobble.AI backend applications leveraging Terraform as IaC. Leveraging the configuration details shared by the Bobble.AI team, Niveus provided equivalent services within GCP, ensuring a smooth transition and continued functionality for users. This AWS to GCP migration resulted in increased scalability, enabled advanced features, and a more robust platform for the Bobble AI Keyboard application.
- Requests from the Android and Web applications are routed to the Application Layer hosted in the Google Kubernetes Engine through a Global Load Balancer
- The Load Balancer was integrated with Google managed Web Firewall (Cloud Armor)
- Static content was stored in Google Cloud Storage and served through the Content Delivery Network (CDN)
- The Application Layer was integrated with the associated Persistent Layer containing MYSQL Cloud SQL, Enterprise Redis for caching, ElasticSearch set up in GKE, and Cloud Storage hosted in GCP
- Managed Prometheus was set up in GCP for event monitoring and alerting
- The migration of databases was achieved by migration of Aurora MySQL to GCP Cloud SQL in staging and production environment and migration of object storage from object storage in AWS S3 to GCP Cloud Storage
- A dedicated project was set up to host and run Gitlab runners in GKE, which was connected to Gitlab for CI/CD jobs. This setup was used to continuously build and deploy microservices to GKE clusters
Key Value Delivered
- Enabled collection of logs from all applications at a central location using Google Cloud Logging for increased observability and reduced MTTR.
- Bobble AI has been able to capture and monitor metrics for their infra and application at a much granular level. This has led to better alerting and usage analysis which has increased availability and decreased cost.
- Seamlessly migrated Bobble AI backend application’s entire workload from AWS to GCP without any major outages or issues
- The entire project team has received accolades and appreciation from the client post the cutover including special callouts from the Google team
Customer Feedback
Technology Stack
Google Kubernetes Engine
Cloud Load Balancing
Content Delivery Network
Google Cloud Storage
Memory Store
Prometheus
Cloud Functions
Cloud Pub/Sub
