Homepage image

DevOps implementation


Problem overview

BlueDraft had developed a Forecasting and Financial Analytics tool for large companies customized by client, which required complex, slow and expensive maintenance and updating work.

Solution overview

We redesigned the application's architecture, simplifying the maintenance, the updating and the management of each client's customizations. At the same time, we standardized the different environments and we automated the deployment in production, significantly reducing the time-to-market, human intervention and error occurrences.

Detailed solution

We began by providing training to introduce good practices and standardize important concepts. Then, we carried out an analysis of the software architecture, developed in Django and React, from which it emerged that the architecture was not scalable and that it wouldn’t be easy to maintain each client’s customizations. We formed two work teams, one focused on the backend and the other on the frontend. As a result, we modified the application’s architecture and we generated automation tools to simplify development management, incorporating, in turn, practices such as continuous integration and continuous delivery.

We redefined and improved the container architecture, and with that done, we installed a demo server using Amazon EC2, Amazon EFS, and Docker. Finally, and in order to simplify the client’s work, we created a development environment based entirely on Docker to minimize the time required to create said environment and to standardize the tools and versions used by the team. As a result, the development environment ended up being identical to the production environment.