Loading...
Case Study
Cleq powering Calmie Toys' e-commerce and content platform for Indonesian Gen Z
100%
Infrastructure as Code Coverage
<5 min
CI/CD Build Time
99.9%
Uptime SLA
They understood that for us, downtime isn't just a technical problem, it's a brand problem. Every minute our store is slow during a drop, we lose a customer who might not come back. They got that, and they built something that reflects it. Honestly, it's the kind of infrastructure we should have had from day one.
Yessy Adzanya
Operational Manager ā¢ Calmie Toys
Outcomes
A production-grade cloud infrastructure that lets Calmie Toys ship fast, survive viral traffic spikes, and scale confidently as their community grows across Indonesia and beyond.
1
Spike-Ready Uptime
Auto-scaling hosting architecture absorbs viral traffic surges without downtime, critical for limited drops and influencer-driven campaigns that generate sudden demand.
2
Database Responses
Optimized schema, connection pooling, and read replica configuration keep product inventory and order data fast under concurrent load across channels.
3
Full Environment Parity
Dev, staging, and production environments are infrastructure-as-code defined, eliminating configuration drift and protecting live storefronts from breaking changes.
The Situation
Calmie Toys is an Indonesian kawaii collectible brand targeting Gen Z through a multi-channel social commerce strategy spanning TikTok, Instagram, Shopee, and Tokopedia. With a rapidly growing community built around limited-edition drops, influencer campaigns, and viral content formats, the brand needed a cloud infrastructure capable of handling unpredictable traffic spikes ā particularly during tentpole moments like Harbolnas (12.12), Singles' Day, and K-pop comeback launches that drive sudden surges in product demand and social engagement.
Calmie's digital presence required always-on reliability across their storefront backend, content delivery, and community-facing touchpoints. With no existing cloud foundation in place and an aggressive content calendar publishing up to 5 formats per week across platforms, the team needed a production-ready environment that could scale with their growth without slowing down their creative and commercial velocity.
Core Challenges
-
Traffic volatility tied to viral moments. Calmie's audience is highly responsive to trends, influencer posts, and limited drops. A single viral TikTok or Shopee flash sale could spike backend load several times above baseline, with no warning. The infrastructure needed to absorb these surges without degraded response times or checkout failures.
-
No cloud architecture in place. The team had a brand, a content strategy, and active social channels, but no defined hosting architecture, environment strategy, or deployment workflow. Everything needed to be built from zero with speed and stability as equal priorities.
-
Multi-environment consistency. With multiple content creators and developers pushing updates frequently, the absence of isolated dev, staging, and production environments created high risk of breaking changes reaching live product pages during peak traffic windows.
-
Database performance under concurrent load. Product inventory lookups, order processing, and user session data needed to stay fast even during simultaneous purchase events across Shopee-linked and direct storefront channels. Without proper indexing and connection pooling, the database would become the bottleneck at exactly the wrong moments.
-
Release risk on a fast-moving content calendar. With new product drops, limited editions, and campaign landing pages shipping weekly, manual deployments were too slow and error-prone. A broken release during a Harbolnas campaign window could directly cost revenue and brand trust.
What We Delivered
End-to-end cloud and DevOps infrastructure for Calmie Toys' storefront backend, database layer, and deployment pipeline.
App Infrastructure & Hosting
Provisioned and configured cloud hosting for Calmie Toys' backend API and storefront services. Architected for high availability with load balancing, auto-scaling policies, and environment isolation ā designed specifically to handle the volatile, influencer-driven traffic patterns of Indonesian Gen Z e-commerce.
CI/CD Pipeline
Designed and implemented a full CI/CD pipeline covering automated testing, build, and deployment stages across dev, staging, and production. Branch-based deployment triggers and Slack notifications give the team full visibility into every release ā enabling Calmie's fast-moving content and product calendar to ship safely and consistently.
Database Architecture & Scaling
Architected the production database layer including schema design review, indexing strategy, connection pooling configuration, and read replica setup. Automated backup schedules and point-in-time recovery protect transactional product and order data. Tuned specifically for the concurrent inventory lookup and purchase patterns that occur during flash sales and limited-edition drops.
How we did it?
We started by mapping Calmie Toys' traffic risk profile ā a brand whose demand is directly tied to viral social moments, influencer posts, and time-boxed tentpole events. Unlike steady-state platforms, Calmie needed infrastructure that could sit quietly at low load most of the time, then absorb a ten-fold spike the moment a TikTok goes viral or a Harbolnas flash sale goes live. This shaped every architectural decision we made.
All infrastructure was defined as code from day one, enabling consistent, repeatable environment provisioning across dev, staging, and production. This eliminated the configuration drift that had previously made deployments risky, and gave the engineering team confidence to ship frequently without fear of breaking the live storefront.
The CI/CD pipeline was built around Calmie's existing Git workflow, with automated test gates on every merge and structured deployment triggers per environment. What had previously been a slow, manual process became a sub-5-minute automated release cycle ā critical for a team publishing new product drops, campaign landing pages, and content-linked storefronts on a weekly cadence.
For the database layer, we profiled the query patterns most likely to cause bottlenecks under load: concurrent inventory reads during drops, order creation during flash sales, and session lookups during high-traffic content moments. Schema optimization, targeted indexing, connection pool tuning, and a read replica for reporting queries brought transaction latency consistently under 100ms even at peak.
The result is a cloud foundation that moves as fast as Calmie's creative team ā reliable enough to protect revenue during their biggest sales moments, and flexible enough to support their long-term expansion from Indonesia's Gen Z market into international kawaii communities.