Renting Colombia optimizes service with Microsoft AI Services

Car rental for personal or business use is a growing trend in the world. For both short-term leases and contracts up to three years, this model has become a very attractive option for many people and businesses who need vehicles, but not the inconvenience that owning them may entail. Its benefits are evident: variable rental periods, elimination of paperwork, and additional cost savings on taxes and maintenance. 

For 23 years, Renting Colombia has been the leader of these services in the country and has expanded its operations quickly. It has more than 35,000 passenger and cargo vehicles and four regional headquarters where more than 500 employees work.   

One of their greatest challenges is to ensure that their customers have complete availability of their vehicles, i.e., that they are in optimal conditions of use. Efficient and timely maintenance is essential for this. With an operation at Renting's scale, it was urgent to implement a technology-based solution that would allow it to transform manual actions into agile automated processes powered by machine learning and AI to identify vehicle service needs. This would allow, for example, the removal of thousands of manual service order approvals and authorizations from each maintenance entry. This process involved "downtime" hours from vehicles and could take up to four hours. Since there were an average of 27,000 service orders with over 100,000 parts and services per month, it was clearly a critical business process.  

With customer satisfaction in the spotlight, the Colombian company identified in Microsoft technology a scalable, secure solution that, in addition to minimizing vehicle out-of-service times, also offered full visibility into the supplies they place in the vehicle for each service order. This must be calculated and guaranteed prior to the vehicle's arrival at the workshop to minimize wait times and to plan variables such as price and frequency of service. 

"Caja Negra": a cloud automation and artificial intelligence solution 

The implementation of the solution called "Caja Negra" adopted Azure AI Services to develop a virtual assistant using machine learning. 

"This solution not only automates orders, but also learns from the past. It identifies, for example, whether the battery has already reached its useful life, the price of a replacement, or in which region of the country it is most common for brake wear to be greater, depending on the topographical conditions," says Daniel García, assistant manager of the Renting Intelligence Center. He adds that it also "allows us to establish usage and maintenance patterns, compare offers and availability of suppliers, streamline and give greater transparency to a process that's discretionary, cumbersome, and subject to many errors."

In turn, Silvia Toro, Manager of Technology Development at Renting Colombia, explains that although there was "a systematized exchange of information with our suppliers, it took time to approve or reject decisions regarding vehicles, and this was something we wanted to improve with Caja Negra. The manual process already had weaknesses because the information was scattered and was based on the different criteria of each supplier. The model was powered by historical information and allows for the resolution of one of the large bottlenecks of an operation such as the Renting one."   

Currently, the virtual assistant automatically and instantly approves about 65% of service orders with an approval or rejection response for 100% of the inputs that make them up. Using Azure microservices, such as Azure Bot Service, Caja Negra performs a statistical evaluation of the data and transforms it into tools so that the virtual assistant is able to reliably respond to the elements that are part of the vehicle's maintenance. This accelerates a process which was previously done by consultants who relied on their own experience and their best judgment. Additionally, the information was slow to arrive and was not standardized. Now Renting experts are dedicated to more strategic tasks for the company and are supported by dashboards generated in Power BI. Here they can find the precise data about approvals made by the virtual assistant and can intervene when decisions require that expert eye that only comes from experience. 

Astrid Correa, Managing Director at Renting Colombia, points out that, at first, automation raised concerns about myths around a possible elimination of jobs and duties. "We had to work on transforming people, not just processes," she insisted. But the concern didn't last long: the model drove such a growth in business volume that Renting employees’ roles became more strategic and less operational. The company also managed to train its staff and build trust in technology, which is now the best partner in its operations. 

Process optimization and increased satisfaction 

Thanks to the adoption of this solution, Renting Colombia was able to reduce the analysis time of service orders by 92%. It went from a process of up to four hours to an automatic process of a few seconds. That's the analysis time the virtual assistant takes.   

Sergio Parra, Head of IT Solutions Technology Operations at Renting Colombia, explains that "the assistant takes only 10 seconds to approve or reject the supplies since it is based on the historical data of a fleet, which can reach 50,000 vehicles."  

This implementation made it possible to optimize the procedure for customers. After all, the most important thing is that every Renting customer can depend on their vehicle all the time and can trust that it is in perfect condition. Following the implementation of Caja Negra, satisfaction levels increased positively not only for the end customer, but also in the internal structure of Renting Colombia. This was also the case for suppliers, as they now have more controls and new levels of security. 

"We are reaching new markets and going through a process of massification. This is a project that drives sustainable fleet growth and the creation of new business models. In this journey, our product is accessible to many more people, so technology is very supportive of our growth, and we are working to stabilize it in our operations," says Silvia Toro, Technology Development Manager.  

And as is often the case in transformation processes, innovation has inspired them to continue along this path. Renting has already implemented a digital app for users that allows each customer to self-manage all their needs, ranging from paying their monthly rent for their vehicle to asking for assistance or monitoring the status of the service order in real time.  The next steps in this journey to digitization are already clear, and the organization is studying analytics models for predictive maintenance which allow it to determine which car requires service before this need becomes apparent.