Car RentalReal Time Analytics

The customer

Our client is a leading company in the car rental sector that operates throughout Italy and Eastern Europe, targeting business and private customers.

The state of the art

Over the years, the company has entered into agreements with different brokers for the management of quotes and car reservations. Specifically, the data on quotes, which do not always translate into reservations, are stored on a relational database. The company therefore expressed the following critical issues:

  • The data remained unused and had duplicate records;
  • It was difficult to integrate data from different sources;
  • Monitoring and visualization would have required long development times and high costs if made with the technologies used by the company.

The challenge

The company therefore turned to Koros Consulting to develop an integrated solution for real-time data analysis through the creation of a pipeline and a dashboard that would allow it to monitor the performance of some KPIs. These are the needs:

  • analyze in real time the progress of estimates;
  • disambiguate duplicate quotes;
  • compare quotes with actual bookings;
  • view the data through dedicated dashboards;
  • carry out historical analyzes.

The solution

To meet the requirements, it was therefore decided to build a real-time data pipeline using the following technologies:

  • Confluent with which, through Kafka Connect, the data are taken in real time from the source databases, then they are disambigued and aggregated;
  • RethinkDB, NoSQL database where Confluent reprocessed data is stored for persistence. The chosen database guarantees efficiency for real-time applications and is natively distributable and scalable;
  • Python Dash for the development of dashboards that allow analysts to view, in self-service mode and in real time, the performance of identified KPIs. It is also possible to carry out, as required, customized historical analyzes using filters and export the data for subsequent analysis using spreadsheets;
  • Docker for on-premise architecture development.

The benefits

The solution created in light of the customer's needs has made it possible to obtain not only performance improvements but above all a precious source of forecasts to support corporate decision-making processes.

Speed

In a few seconds, after their storage on the different sources, the data are integrated and made available. The use of Docker has also made it possible to significantly reduce the hardware and software infrastructure configuration times.

Integrity

Thanks to the elimination of duplicates, the starting dataset was made consistent and the anomalies contained therein were removed.

Insights

Using the data visualization tools, it is possible to monitor KPIs in real time and extract information from the data to make “data-driven” decisions.