Loading...
Deployment Frequency
Environments Managed
CI/CD Pipeline Build Time

Working with the team completely changed how we handle our cloud infrastructure. Before, we were dealing with slow deployments and unpredictable costs, especially as our refill station network across Bali kept growing. After the optimization, everything became more streamlined faster load times, better scalability, and significantly lower overhead. It’s allowed us to focus on expanding EZ Water’s mission of reducing plastic waste instead of constantly worrying about backend performance.

Director • EZ Water
A production-grade cloud infrastructure that lets EZ Water ship fast, scale confidently, and operate reliably across all partners and app touchpoints.
We plan before we touch anything, and we do not cut over until we are certain nothing will break. Let us take a look at your setup.
Automated CI/CD pipelines eliminated manual releases, reducing deployment time from hours to under 5 minutes with rollback capability.
Highly available hosting architecture with health checks and auto-recovery ensures kiosk dispensing and app sessions are never interrupted.
Optimized schema, connection pooling, and read replica configuration handle concurrent refill sessions and real-time balance updates without degradation.
Unified logging and alerting pipeline gives the team real-time visibility into app errors, database load, and infrastructure health.
EZ Water was building Bali's next-generation smart water refill network — a system spanning outdoor Smart ATM kiosks, home RO units, and a consumer-facing NFC-powered mobile app with a digital Water Wallet. The platform required always-on, resilient infrastructure capable of handling real-time dispensing events, user authentication, prepaid liter balances, and loyalty logic — all running reliably in a tropical environment with variable connectivity. With a hard launch deadline tied to Bali's 2026 single-use plastic bottle ban, the engineering team needed a production-ready cloud environment fast, with zero tolerance for downtime at the kiosk level.
Cleq identified several challenges that were limiting scalability and increasing operational complexity:
Synchronizing Water Devices with the Cloud Refill stations and home units operate in environments with unstable connectivity and power. The system must handle offline scenarios, delayed data, and synchronization issues while maintaining consistent operations.
Ensuring Reliable, Real-Time Data The platform depends on accurate water quality and machine status data. The challenge is turning noisy, asynchronous sensor inputs into a trustworthy source of truth, where correctness is more critical than speed.
Handling Multi-System Complexity The solution combines consumer apps, operational tools, and logistics systems. Each has different scaling and reliability needs, requiring a flexible and well-structured architecture.
End-to-end cloud and DevOps infrastructure for EZ Water's app, database, and deployment pipeline.
App Infrastructure & Hosting
Provisioned and configured cloud hosting for EZ Water's backend API and mobile app services. Architected for high availability with load balancing, environment isolation, and auto-scaling policies suited to variable Bali traffic patterns.
CI/CD Pipeline
Designed and implemented a full CI/CD pipeline covering automated testing, build, and deployment stages across dev, staging, and production. Integrated branch-based deployment triggers and Slack notifications for release visibility across the engineering team.
Database Architecture & Scaling
Architected the production database layer including schema design review, indexing strategy, connection pooling configuration, and read replica setup. Established automated backup schedules and point-in-time recovery to protect transactional refill and wallet data.
We began with an infrastructure audit of EZ Water's existing codebase and deployment assumptions, then designed a cloud architecture tailored to their three core runtime concerns: the backend API serving the mobile app, the real-time database handling refill sessions and wallet balances, and the deployment pipeline connecting their engineering workflow to production.
All infrastructure was defined as code from day one, enabling repeatable environment provisioning and eliminating configuration drift between dev, staging, and production. The CI/CD pipeline was built around their existing Git branching strategy, with automated test runs gating every merge to staging and production — cutting release cycles from ad hoc manual pushes to structured, auditable deployments.
For the database layer, we profiled query patterns specific to EZ Water's domain — concurrent liter deductions, refill event logging, loyalty streak calculations — and tuned the schema, indexes, and connection pool sizing accordingly. A read replica was introduced to offload reporting queries from the primary write path, keeping transaction latency consistently under 100ms.
The result is an infrastructure stack that EZ Water's team can operate confidently as they expand kiosk locations across Bali and eventually into international franchise markets.