Fully Data-Driven decision making platform

Savia was born as a start-up created by MAPFRE, one of the major insurance companies in Spain with presence in more than 40 countries. Savia is a single digital health product with an internal MAPFRE team forming the new business. It’s a 100% digital company that offers to the users, free of charge, the possibility of chatting with the specialists in its medical directory, among other health digital services as well as a marketplace of physical health assistance products.

Savia was conceived as a solution to resolve medical consultations without the need for long waits, both online and in person. All services are created with the intention of bringing quality medical and health care closer to all types of people and families establishing a relationship of total trust. 

As an example of its ability to respond quickly to medical consultations, with the emergence of the Covid-19 pandemic, Savia launched a service that allowed the user to consult the doctors about the appearance of possible symptoms associated with Covid-19.

Digital Transformation: The Key Challenge for Companies in the Post-Pandemic Era.

For a few years, and especially since the pandemic, we find ourselves in a scenario in which all companies are focusing their efforts and resources on digitization.

Companies need new tools, technology and digital ecosystems to provide greater value to customers, new experiences, solutions and business models. The ability to offer all these things to the user in a quick and efficient way becomes a determining factor in positioning companies in the market.

Savia needed a fully data-driven decision-making platform. The company created a data platform that was initially being built in GCP with the approach that it would only be a healthcare start-up. After presenting to MAPFRE this platform, Savia got the challenge to evolve its platform to a multi-tenant one in order to support other MAPFRE digital services that already had their own legacy systems.

The main challenge here was fulfilling the business needs of:

  • Migration from a Django+Celery approach (adhoc for each tenant and with poor performance), to services decoupling and Cloud based ETL solution. 
  • Flexibility: ETLs both in streaming and batch mode.
  • EDA (Event Driven Architecture): Being able to react when files are created in buckets, Streaming events are coming from mobile apps and websites, or scheduled tasks.
  • Scalability needs as files and events are increasing.
  • Data observability. Event sourcing approach in order to be able to replay events at any point in time.
Solution on Google Cloud Platform

Upon Keepler’s entry into the project, Savia’s data platform is called SDM and has already grown, so it is presented to MAPFRE, which decides to support it and convert this platform into a multi-tenant platform for other MAPFRE digital businesses, such as the loyalty card, the self-service health service and the Home non-insurance service. These three business verticals are introduced on this platform. Keepler’s work was making a solid data lake that could make all the information available for the different business units.

This is achieved through a three-layered system consisting of: a raw data layer; a data processing layer and a layer that connects to the dashboard. This dashboard turns this tool into something that is not exactly a reporting platform, but also provides the different business units with dynamic and customizable dashboards to obtain the information “on the fly”.

The business needs were covered.

The Google Cloud Platform services used were as follows:

  • Cloud DNS: Endpoints for mapper files
  • Cloud Functions: Specific Tasks that only run when invoked
  • GKE: Connector Manager to Data Sources
  • Cloud Load Balancing: Load Balancing for endpoints
  • Virtual Private Cloud: VPCs
  • Pub/Sub: Messaging system
  • Dataflow: Data Pipelines (Apache Beam)
  • BigQuery: Data Warehouse
  • Looker: Data Visualization
  • Cloud Build: Listens to Bitbucket and deploys Terraform infrastructure
  • Artifact Registry: Hosting of Docker images and private Phyton repos
  • Cloud IAM: Permissions’ Management
  • Cloud Data Loss Prevention: Automatically Masks sensitive data
  • Secret Manager: Credentials Manager
  • Cloud Logging
  • Cloud Monitoring: Supervision / Monitoring / Logging
  • Error Reporting
  • Cloud SQL: Event-Sourcing Data Traceability
  • Savia’s main goal was to achieve a fully data-driven decision-making platform. The solution implemented by Keepler made this possible. Now Savia has a solid data lake that makes all the information available to the different business units. This positions the business as a company that is taking unique actions in the market.

  • Keepler provides data analysis resources with expertise in Looker, a profile very scarce in the market, positioning Savia as a pioneer company. 

  • Keepler is currently working on parallelizing the transformation of rows for big files (1-4GB) in Dataflow. This will reduce both processing and debugging time optimizing the experience and making it more efficient.

Keepler is a full-stack analytics services company specialized in the design, construction, deployment and operation of advanced public cloud analytics custom-made solutions. We bring to the market the Data Product concept, which is a fully automated, public cloud services-based, tailored software that adds advanced analytics, data engineering, massive data processing, and monitoring features. In addition, we help our customers transition to using public cloud services securely and improve data governance to make the organization more data-centric.

Let’s talk!

If you want to know more or if you want us to develop a proposal for your specific use, contact us and we’ll talk.