Case study · Rental and mobility · Streaming

Real-time rental pricing on Confluent Kafka for a leading auto rental firm

A leading auto rental firm replaced a traditional pricing model with a streaming platform that adjusts rental prices as availability and demand change.

Industry
Auto rental
Capability
Real-time streaming
Platform
Confluent Kafka
Related POD
Kafka POD

Context

The client is a leading auto rental firm. It wanted to price rentals from real-time data so that it could increase revenue and improve customer satisfaction, and it engaged VerticalServe to build the pricing platform on Confluent Kafka.

The problem

The firm priced rentals with a traditional model that did not account well for fluctuations in demand. The result was missed revenue opportunities and dissatisfied customers.

What was built

VerticalServe designed and implemented a real-time pricing platform on Confluent Kafka with six parts.

Flow: reservation and vehicle tracking systems feed availability data into Confluent Kafka, where Kafka Streams and the existing pricing services produce prices that Kafka Connect exports to customer-facing applications and internal reporting. SOURCES Reservation systems Vehicle tracking systems CONFLUENT KAFKA Availability data Kafka Streams pricing Existing pricing services KAFKA CONNECT EXPORT Customer-facing applications Internal reporting High availability · fault tolerance · monitoring

Flow: reservation and vehicle tracking systems feed availability data into Confluent Kafka, where Kafka Streams and the existing pricing services produce prices that Kafka Connect exports to customer-facing applications and internal reporting.

  • Availability data ingestion: real-time availability data is ingested from sources such as reservation systems and vehicle tracking systems.
  • Pricing services integration: the platform is integrated with the firm’s existing pricing services, so prices can be adjusted from real-time data.
  • Kafka Streams: data is processed and analysed in real time to produce pricing recommendations from demand, vehicle availability and market conditions.
  • Kafka Connect for export: optimized pricing data is exported to customer-facing applications and internal reporting tools.
  • High availability: a highly available, fault-tolerant infrastructure keeps the platform running.
  • Monitoring: monitoring tools track the platform’s performance and identify issues.

Stack

Confluent Kafka Kafka Streams Kafka Connect Monitoring

How it was delivered

The team first set up the Confluent Kafka infrastructure and the ingestion of real-time availability data, configured so the data is processed and stored efficiently and is ready for analysis.

Next came integration with the firm’s existing pricing services, followed by the Kafka Streams processing that turns demand, availability and market conditions into pricing recommendations.

Kafka Connect was then added to export the optimized pricing data to customer-facing applications and internal reporting tools, so the people and systems that need a price always see the current one.

Finally the team put the highly available, fault-tolerant infrastructure in place and integrated monitoring.

The result

  • Dynamic pricing: rental prices now adjust from real-time data, in support of revenue and customer satisfaction.
  • Pricing recommendations are always based on the latest information, because the data is processed in real time.
  • Optimized prices reach customer-facing applications and reporting tools efficiently through Kafka Connect.
  • The fault-tolerant infrastructure keeps the platform in continuous operation and limits downtime.
  • Monitoring lets the firm identify and address issues more effectively.

The original write-up reports these outcomes without figures.

What it means for a similar team

  • If prices are set by a model that cannot see demand as it moves, start with the availability feed. Everything else in this platform depends on it.
  • Existing pricing services do not have to be replaced. Here they were integrated and fed with real-time data.
  • Treat high availability and monitoring as part of the build. Pricing sits on the revenue path, so downtime costs money.

This case study was first published on our Medium publication. Original article

Engage

Start a similar POD

Tell us the outcome you want and your constraints. We will reply with a discovery proposal.

Talk to us

Working on something similar?

An Innovation POD is a dedicated team that builds inside your environment. Tell us what you need to ship.

Start a POD