ManoMano: Image Microservice & Import MVP
ManoMano is a European marketplace focused on DIY and gardening, offering one of the largest selections of tools, materials and branded products at competitive prices.
To continue scaling its platform and improving performance, ManoMano was searching for a technological partner capable of implementing new features, enhancing existing components and bringing best development practices into the team. This included the creation of an image management microservice and the completion of a critical MVP for product importing.
One of the key challenges was developing an image management system capable of running on a server with only 256 MB of RAM, requiring an algorithm capable of processing images incrementally instead of loading them entirely in memory. Additionally, the image processing workflow was blocking the product import flow, demanding optimal performance and minimal latency. At the same time, the product import MVP was inherited from a large legacy project, which required a steep learning curve and extensive refactoring to finalize within deadlines.
Optimizing image processing and completing the product import system
ManoMano looked for a technological partner capable of developing new platform features and improving existing software. The first key objective was implementing an image management microservice capable of storing images, validating them, and providing access through dynamic rescaling.
Once the microservice was delivered, we collaborated with ManoMano’s internal team to finalize the MVP of their product import system, a critical component of their marketplace operations.
The microservice had to run on a server with only 256 MB of RAM. To achieve this, we designed an algorithm capable of processing images in small chunks rather than loading them fully in memory, enabling high performance under tight constraints.
We also implemented an AWS Lambda that consumed S3 images and enabled dynamic rescaling through URL filters, improving flexibility and efficiency for multiple use cases.
Technologies Used
Technologies used: Kotlin, Java.
Results
- The image processing microservice was delivered two weeks ahead of schedule and operated smoothly from day one.
- The Lambda-based dynamic rescaling system improved flexibility and performance across the platform.
- The product import MVP was successfully completed and deployed to production within the expected timeline.
- Refactoring and best practices improved maintainability and reduced technical debt of the inherited project.
We are ready for new challenges
