Introducing Streamhouse: the open data architecture for AI | Learn More
Real-Time Spam Detection
Use data streaming for artificial intelligence (AI) and machine learning (ML)-based anomaly detection to automate spam blocking. Stay ahead of spammers to ensure customer trust and reduce network load and OPEX costs.
Unleash the Power of AI to Combat Spam
As SMS spammers become increasingly sophisticated and find ways of circumventing rules-based approaches, it’s time to leverage AI and ML to combat spam in real time. To do this, AI needs as much data as possible – the more up-to-the-minute and enriched data there is, the better spam detection will be.
Leverage Confluent’s data streaming platform to improve ML model training by bringing together data from any source and replacing batch processing with stream processing. Continuously ingest, transform, and use real-time data to calculate an anomaly score and automate spam detection and blocking.
Block spammers in seconds instead of hours, making it unprofitable for them to continue their attacks.
Improve customer experiences by reducing unsolicited messages.
Save operational costs with reduced network load from less non-value add traffic.
Build with Confluent
This use case leverages the following building blocks in Confluent Cloud.
Reference Architecture
Eliminate batch data and allow SMSC to stream records to Confluent as they are created in real time. Continuously capture changes to customer data held in relational databases such as Oracle by using the Oracle CDC Source Connector to write to topics in Confluent Cloud. Client-side field-level encryption protects any sensitive customer data.
Continuously transform and enrich data from any source using Flink stream processing in Confluent. Calculate metrics in real time (e.g., numbers of messages, message size) based on a specified time window. This enables the application running the ML model to evaluate all features provided with the metrics, and generate an anomaly detection score. Based on that, a score-handling microservice determines if a specific customer should be blocked or not and calls the blocking API.
Ensure data is secure, compliant, high quality and ready to use with Stream Governance. Easily tag and add business metadata to make real-time data products easily discoverable and accessible for the rest of the organization and for any AI model to use.