Kaaja: the first platform in Italy for private online real estate auctions
The Kaaja digital platform has one goal: to simplify the process of buying and selling real estate. It is unique in its kind because, thanks to proprietary algorithms, it digitises all intermediation operations, from promotion to offer, up to the registration of the compromise, entirely notarised on blockchain. Which means shorter times - it concludes no later than six weeks - and transparency from the first moment.
The Challenge
We had the opportunity to get to know the internal team and the project in 2020 through a technical discovery activity aimed at punctually exploring the different features and ways to build the platform, also supported by FifthBeat, partner for UX/UI design.
Following the initial discovery phase, the Claranet and FifthBeat teams began collaborating with the Kaaja team on every aspect: UX, UI, frontend, backend, and ops.
We agreed to work with a Satisfied or Refunded contractual modality (with weekly iterations) with the aim of releasing value right away.
In a few weeks we put online, in a protected domain, the first features such as registration and login, created through the AWS Cognito system. This AWS service allowed us to implement the registration and authentication flow in a secure and tested way, leaving the team free to focus on the business logic that brings the most value to the platform.
Led by the Kaaja team, the FifthBeat team focused on the UX by creating both mobile and desktop designs; these inputs, along with the output of technology spike activities for the more complex backlog items, helped define the scope of each release, prioritising releases based on customer guidance.
The solution
The most modern frontend and backend technologies were adopted for the development of the platform, adopting the CQRS + EventSourcing architecture for the latter. For content management we opted for a headless CMS, which allowed us to speed up the insertion of content (real estate and auctions) into the platform. Furthermore, a custom backoffice was created to manage auctions autonomously and with domain logics that would have been more difficult to integrate into the CMS.
Every week with the Kaaja team we agreed on the various features to be developed in order of priority and value for the business.
Regarding the Cloud infrastructure, it was designed by identifying the most elastic and scalable solutions that AWS allows to adopt, thus creating a resilient and multi-level infrastructure.
The project is composed of three different AWS environments and accounts to ensure maximum isolation in terms of security. The services used for this purpose are multiple (more than twenty) given the complexity of the project: the back office uses CloudFront and S3 for the type of framework present, while for the APIs the principle of decoupling services was adopted using serverless applications on Lambda and API Gateway. For the application core, however, we used the ECS service – with EC2 – ensuring the distribution of microservices on different availability zones. To complete the use of ECS we configured the EventBridge and Step Functions services for the execution of periodic tasks.

The result
The goal of the first release was to allow customers to purchase a house through auction, after registration. We achieved it in about twelve weekly iterations: a very challenging and demanding goal, also achieved thanks to the careful selection of the most valuable user stories for that release together with FifthBeat and under the guidance of the Kaaja team.
After the first release, we listened to the new priorities of the Kaaja team and agreed with FifthBeat on the content of the next iterations, improving the iterative process through internal retrospectives.
Latest articles

Global satellite communications company gets a bird's-eye view of data migration to AWS

Cerved improves data quality and reduces costs with serverless AWS machine-learning

Sailsquare: migrate to cloud by optimising infrastructure

Daldoss Elevetronic: Cloud Innovation to Optimise Sales Work

Scattered Spider don’t break in. They log in.