The Thread Lightly product service is a reconstructed API for a mock e-commerce clothing retailer. The goal was to optimize the back end for speed and responsiveness under heavier loads. The project consisted of working with an inherited front-end and making it work with the same endpoints and formats the client-side code was expecting. The work was carried out within a team of three, each of whom was responsible for a different API service. I was responsible for the products service.
This project is a sister project to Thread Lightly retail portal.
- Composed ELT pipeline using MongoDB Compass, the Mongo aggregation pipeline, and the Mongo CLI to seed database with over 22 million records.
- Improved query retrieval speeds by 97% over the legacy code by taking advantage of indexing and embedded documents, decreasing response time to as little as 4ms.
- Maintained a 0% error rate at 1000 requests per second during stress testing.
- Node.js
- Express.js
- MongoDB
- Mongoose
- MongoDB Compass
- AWS
- Artillery
- Docker
- Alex Hernandez
- Addie Johnsen
- Eric Handley