Marketplace vendor platform

Our client was looking for a trusted IT partner that could develop a platform for marketplace owners. The main purpose was to have one place where vendors could see all the sales data of customers – which goods sell the best, which warehouses need more things to send, and which ones have a full inventory. It was supposed to be simple to use, as vendors from all around the world should be able to easily browse it. After establishing the project’s shared vision, our team was entrusted with the task of designing and developing a platform.

The most difficult part during the project development was integrating with our client’s e-commerce ERP system. It was already utilized daily, and all order information was stored there. But we managed to overcome the difficulties! We created several unique solutions that allowed us to extract the data and put it on the platform in real time. 

Despite the remote collaboration, the workflow went flawlessly. Together we kept up to date on the project’s status and tasks daily. We both communicated using Slack and emails, which allowed us to work with each other regardless of distance.

The idea of the platform was to sell items in the most popular local marketplaces quickly, easily, and efficiently. To overcome difficulties with an e-commerce ERP system, we had to develop the integration from scratch. We couldn’t use available libraries or frameworks as they were limited by API communication and couldn’t meet the client’s expectations. Our client stipulated that data has to be easily accessible and analyzed, such as which channel is the most successful for a given vendor, what things sell best, and whether there are channels that are better for selling low-priced or high-priced products, to assess order value. 

We looked carefully into these needs and managed to deliver the expected technological solutions. The technology we used was Vue.js for front-end development and Ruby on Rails for the back-end side. With a custom wrapper based on Sidekiq API, we could pool new records easily. We used the Redis database and, after some trial and error, we managed to minify the JSON payloads of individual orders to speed up the platform. It resulted in a drop from tens of seconds to a few hundred milliseconds. By the end of the project, we had extra knowledge for further development!