Client & context
A large online hypermarket selling hundreds of thousands of products was running on a heavily customized in-house platform. Over time, the system had become:
- Difficult to maintain – legacy code, lack of documentation, tight coupling everywhere.
- Risky to change – fear of breaking core functionality.
- Under growing pressure from traffic growth and new requirements.
I joined as the sole developer responsible for the core platform.
---
Challenges
- High load: The platform was serving over 3 million requests per day and handling 350,000+ products.
- Legacy stack: Much of the system was built years ago, with ad-hoc changes and inconsistent patterns.
- Low robustness: Failures in one part of the system could easily cascade into others.
- Limited visibility: There were gaps in monitoring, metrics, and logging, making it harder to diagnose issues.
The business needed to keep the platform running and evolving without a full rewrite or major downtime.
---
What I did
1. Deep research & stabilization
- Performed a quick architecture and code review to identify the most fragile areas.
- Added basic monitoring and logging in key places to understand traffic patterns and failure modes.
- Implemented targeted bug fixes and small refactors to stabilize the most critical paths.
2. Performance & reliability improvements
- Optimized database queries and added appropriate indexes for high-traffic operations.
- Improved caching strategy for frequently accessed data (product information, categories, etc.).
- Tuned web server and PHP settings to match real-world usage patterns.
- Improved handling of long-running tasks and background jobs.
3. Progressive modernization
- Introduced more consistent architectural patterns without disrupting the whole system.
- Updated parts of the stack to more modern versions where it was safe and valuable.
- Added or improved internal tools to make maintenance and feature development easier.
4. Close collaboration with the business
- Worked closely with business stakeholders to prioritize changes that had the highest impact on stability and customer experience.
- Provided clear explanations of risk vs. benefit for larger refactors and upgrades.
- Helped evolve the roadmap so that technical improvements were aligned with business goals.
---
Results
- The platform remained stable under high load, serving millions of requests per day.
- Performance improved, with faster responses on critical user paths.
- The system became safer and cheaper to evolve: New features could be added without fear of breaking everything.
- The company avoided a costly and risky full rewrite and continued to operate on a platform that was now robust, better understood, and more maintainable.
---
Tech & responsibilities
- Role: Sole developer / principal engineer for the core platform
- Technologies: Custom PHP/Laravel-based backend, MySQL, web server configuration, caching, background jobs, internal tools
- Scope: Architecture, hands-on coding, performance tuning, debugging, production support, stakeholder communication
---
If you have a high-load legacy platform that you’re afraid to touch—but can’t afford to replace—I can help you stabilize, modernize, and scale it pragmatically.
Back to all case studies
