[Webinar] Bringing Flink to On-Prem and Private Clouds | Register Now
In-Store Personalization
Confluent enables retail brands to stream, govern, and process on- and off-line data from multiple sources in order to provide consumers with highly personalized in-store shopping experiences.
Deliver Tailored In-Store Shopping Experiences
The idea of ‘personalization’ is fundamental to eCommerce brands. Personalized product recommendations, discounts, and advertisements (to name just a few applications) based on the online behavior of consumers have helped fuel the growth of many retail companies. Now, with the help of data streaming, it’s possible to replicate personalized online shopping experiences in physical stores. Consumers’ behavior can be joined across on- and offline channels in real time in order to deliver a tailored in-store experience.
Confluent powers in-store personalization by streaming, processing, and governing data between multiple sources and destinations.
Elastically scale streaming data pipelines to throughputs of GBs/second in order to manage spikes in demand (e.g., Black Friday).
Deliver personalized messages, discounts, and recommendations to consumers while they’re in-store via their mobile devices. This is made possible by <10m/s latencies delivered by Confluent’s Kora engine.
Ensure the high availability of streaming pipelines for in-store experiences with multi-zone and multi-region clusters.
Build with Confluent
This use case leverages the following building blocks in Confluent Cloud.
Reference Architecture
This diagram demonstrates a sample architecture of how to deliver in-store notifications to customers based on previous transactions, web behavior, and their current physical location. It involves the ingestion and processing of data from different sources (e.g., PostgreSQL and Google Cloud Storage bucket) and a custom Ably connector, which enables edge interaction with browsers, mobile devices, and IoT devices.
Event data streamed from mobile device to Confluent via REST proxy. Transaction and Membership loyalty data streamed to Confluent via PostgreSQL CDC connector.
Streaming processing used to connect and filter disparate datasets.
Data streaming to Ably via a custom Ably Sink connector, which pushes notification to customers based on account and physical location.