Real-Time Analytics of Customer Financial Activities with Apache Flink

Explore the emergence of real-time analytics in financial institutions through this conference talk from Philly ETE 2016. Dive into Capital One's experience building a real-time analytics application using Apache Flink for processing customer activities. Learn about the shift from traditional batch analytics to real-time processing, understanding the business drivers, technology enablers, and social trends behind this transformation. Discover how Apache Flink's event processing capabilities and sophisticated windowing semantics make it an ideal candidate for financial use cases. Examine the architecture of a customer activity log system, including implementation details on real-time alerts, data enrichment, transformations, and time-based sliding window aggregates. Gain insights into how this analytics platform can be applied to various use cases such as fraud detection, and understand the spectrum of analytics from batch to real-time processing in modern banking systems.

Intro
1.2 Capital One Technology
2.1 Traditional Batch Analytics
2.3 Limitations of Traditional Batch Analytics
The Great Paradigm Shift - Real-Time Analytics
3.1 What is Fast Data?
3.2 Real-Time v/s Batch - Water Heater
3.3 Real-Time Analytics
3.4 Real-Time Analytics - Use Cases
What are the Drivers?
4.1 Business Drivers
4.2 Technology Drivers
4.3 Social Trends
4.4 New Industries and New Use Cases
5.1 Apache Flink as the Next Generation of Big Data Analytics
Apache Flink Stack
Business Use Case: Customer Activity Event Logs
6.2 Architecture of Customer Activity Logs
6.3 Real-Time Analytics with CAL Data
6.4 Implementation Details
6.4.2 Real-Time Alerts
6.4.3 Enrichment
6.4.4 Transformations
6.4.5 Window Aggregates - Time-based Sliding Window
6.4.5. Real-Time Index and Search
6.5 Generic Pattern Supports A Class of Use Cases
6.6 The Analytics Spectrum - Batch & Real-Time