Assembling Commerce
Fetching retail intelligence layer...
Reducing Cloud Costs by 90-95% with Serverless Modernization
Rava AI ran its GenAI SaaS platform on an inefficient VM-based architecture (6 active GCP Compute Engine instances) for production, staging, Node.js, and LLM services. The setup also suffered from service outages requiring manual interventions, lack of auto-scaling/self-healing, and an unreliable in-memory logging configuration.
We redesigned the entire infrastructure using a serverless-first approach on Google Cloud Platform. Core backend services were migrated to Cloud Run containers, and event-driven workloads (like web scrapers) were converted to run on-demand. Dependency on always-on VMs was eliminated.We enabled auto-scaling to zero when idle, stateless service patterns, and centralized logging via Cloud Logging.
Migrated backend services to Cloud Run, providing automatic scaling, stateless operations, and request-based billing.
Replaced always-on Compute Engine VMs with a serverless setup that scales down to zero when idle, eliminating computing waste.
Refactored scrapers and batch processing tasks to execute on-demand rather than running constantly on background servers.
Achieved a 90-95% overall reduction in cloud compute costs by eliminating idle resource waste.
Enabled seamless auto-scaling to handle traffic spikes dynamically without manual intervention.
Centralized system diagnostics by moving from unreliable in-memory logs to GCP Cloud Logging.
"Brilliant results and exceptional technical execution."
Rava AI is a GenAI-based SaaS platform specializing in advanced content creation, workflow automation, and marketing execution. Website: https://rava.ai/
Book a Similar Project