At Cencosud, I built internal retail applications for warehouse management and item delivery, streamlining operations for over 500 retail locations. I engineered serverless backends with AWS Lambda, implemented Infrastructure as Code with Terraform, and created responsive frontends with React and Gatsby—all while optimizing costs and improving operational efficiency.
Key Contributions
Built Internal Retail Applications
- Problem: Warehouse and delivery operations were manual and inefficient across 500+ retail locations, causing delays in order processing
- Action: Built comprehensive internal applications for warehouse management and item delivery using React/Gatsby for the frontend and serverless Node.js functions for the backend
- Result: Streamlined operations across 500+ retail locations and reduced order processing time by 35%
Engineered Serverless Backend
- Problem: Traditional EC2-based infrastructure was costly, difficult to scale, and required significant maintenance overhead
- Action: Engineered 20+ AWS Lambda functions using Node.js with auto-scaling capabilities, eliminating the need for always-on servers
- Result: Enabled automatic scaling to meet demand and reduced infrastructure costs by 40% compared to traditional deployments
Implemented Infrastructure as Code
- Problem: Manual provisioning and managing 50+ AWS resources across multiple environments was error-prone and time-consuming
- Action: Adopted Terraform for Infrastructure as Code, codifying all AWS resources (Lambdas, DynamoDB, API Gateway)
- Result: Reduced infrastructure provisioning time from days to hours and eliminated configuration drift between environments
Designed Responsive Frontend
- Problem: Warehouse staff needed a reliable, fast interface to manage orders and deliveries across devices
- Action: Built a responsive frontend with React/Gatsby, connected to DynamoDB for real-time data access and optimized query patterns
- Result: Supported 500,000+ daily API requests with sub-100ms response times and improved user experience
- Problem: Generating realistic test data for local DynamoDB instances was manual and inconsistent
- Action: Developed a Go CLI tool that created dummy data for local DynamoDB instances, enabling comprehensive testing of Lambda functions
- Result: Enabled efficient offline testing and reduced integration issues by 45%
Established CI/CD Pipeline
- Problem: Deployments were manual, slow, and prone to human error
- Action: Implemented a full CI/CD pipeline with GitLab, automating testing, building, and deployment processes
- Result: Reduced deployment failures by 50% and accelerated feature delivery
Impact Summary
- 500+ retail locations supported
- 35% faster order processing
- 40% reduction in infrastructure costs
- 20+ AWS Lambda functions deployed
- 500,000+ daily API requests handled
- Sub-100ms response times achieved
- 45% reduction in integration issues
- 50% fewer deployment failures