Date Range:
July - December ( 2024)
Comparison Period:
July - December ( 2023)
- Total Sale49%
- Returning Customer Rate7%
- Orders51%
- New Customers43%
- Sessions67%
- Reached Checkout Rate63%
- Page Views54%
As Nestasia’s operations expanded, so did the strain on their backend systems. We stepped in to re-architect their infrastructure, reduce costs, and ensure stability during high-volume operations while paving the way for sustainable growth.
Here’s How We Approached It
Understanding the Existing Architecture
Nestasia operated on a custom Shopify setup integrated with an internal admin portal that enabled bulk product updates, real-time price syncing, and metadata management. While effective for day-to-day operations, the system wasn’t built to handle large-scale updates without performance strain.
Identifying the Bottlenecks
As the volume of updates grew, especially during bulk CSV uploads, the platform began to experience slowdowns and outages. These issues stemmed from architectural limitations that made it difficult to process high data loads efficiently.
Defining the Core Objectives
Our goal was to stabilize performance during heavy operations, eliminate system downtimes, and bring down monthly infrastructure costs—all without limiting the client’s ability to scale or manage product data flexibly.
The Challenges We Solved
Frequent System Crashes During Bulk Uploads
The client regularly uploaded large CSVs, triggering real-time updates between the admin portal and Shopify. These operations frequently overwhelmed the system.
Webhook Overload Causing CPU Spikes
A two-way sync between Shopify and the admin portal triggered numerous concurrent webhook events. These overwhelmed the server’s CPU, leading to application slowdowns and crashes.
Infrastructure Cost Challenges
The original setup experienced frequent spikes in database IOPS during bulk operations, which led to automatic upscaling of the RDS instance running on Aurora Serverless. This dynamic scaling increased costs significantly, as the system consumed more Aurora Capacity Units (ACUs) during peak utilization even though such high loads occurred only sporadically.
Inefficiency in Scaling
Upscaling the database and application instances to handle peak loads was not a viable long-term solution due to high cost implications.
What We Delivered
Solutions We Delivered
Metrics
65% Reduction in Infrastructure Costs:
Achieved a reduction of nearly 65% in monthly infrastructure costs.
Bulk Upload Performance
The system, which once imposed strict limits on CSV size and record count, now supports unrestricted bulk uploads without performance failures or downtime. Large-scale operations are processed reliably within a timeframe that balances speed with platform stability, meeting the client's operational needs.
Data Sync Handling Efficiency
Incoming data sync requests are now processed in controlled batches, preventing system overload during peak activity and ensuring smooth, uninterrupted operations.
Testimonial
Our Client Review
"Mobikasa exceeded our expectations by optimizing our app performance and streamlining website functionality. Their proactive communication and efficient project management ensured smooth delivery and improved system reliability. We’re very satisfied with the results and ongoing support."
Anurag Agrawal
Founder & Owner
Explore Full Spectrum Solution