Before the start of the project, we held a three-day online Lean Inception workshop, attended by the client, and the RUBICON team where we agreed on the Product vision and Project objectives/requirements. By the end of the project, we managed to develop a digital marketing marketplace with a simple and modern UX/UI design that enables users to find skilled marketers to run online campaigns, measure and define success, and grow flexibly in a cost-effective manner. The project was delivered in phases, where RUBICON first developed the MVP (Minimum Viable Product) and after that the MMP (minimal marketable product).

Due to the project’s complexity, at the very beginning of the project (Sprint 0), the Solution Architect introduced Event Storming to the project team and business. This workshop helped the whole team gain a detailed understanding of the business domain and requirements, clear misunderstandings, and map complex processes. 

Overall, our goal was to build a platform that merges the functions of a services marketplace and a tech-enabled ad consultant and quickly connects brands with the right solution, bringing agility to the procurement process. Also, we had to allow accurate attribution of results of marketing campaigns, measuring performance, alerting for missed opportunities, and influencing future strategy. Another aim was to ensure visibility and control over the client’s marketing activities, giving confidence in the campaign results and subsequent decisions made by the marketing team. Finally, we had to allow brands the flexibility to pay based on their target outcomes.

We created a cloud-native solution using Azure Platform as a Service (PaaS) offering and lowered effort to a minimum so the client wouldn’t have to maintain and manage resources such as storage, servers, databases, applications, services, etc.

Following industry-proven practices for scalable and performant cloud solutions, RUBICON implemented the solution shown in the diagram below:

This solution architecture diagram describes the architecture of the platform:

• Web application built with React/TypeScript that provides all the client’s platform functionalities via a web browser

• Identity provider for authentication and authorization

• Backend REST API implemented using ASP.NET Core WebAPI and Azure App Services.

• Workers perform long-running workflows and scheduled jobs, that are implemented using Azure Function and Azure Durable Functions serverless offering. Backend APIs communicate with the worker through message queues. All long-running workflows and data pipelines are done asynchronously by the worker. The worker is triggered by messages on the queue or runs on a schedule for batch processing. With this approach, backend APIs are decoupled from the worker using asynchronous messaging and both APIs and worker can be scaled independently

• For storage, we choose to use multiple storage technologies (polyglot persistence). The architecture uses Azure SQL Database as a relational database, Azure Cosmos DB as a document database for semi-structured marketing data, and Azure Blob Storage for long-term blob file archive. Data partitioning was done to support improved scalability and improve query performance