Real-Time Financial Position Aggregation Across Customer Accounts

The patent segments the monolithic banking data processing system into multiple specialized microservices including event stream compute devices, position determination compute devices, and various financial product services. Each microservice handles specific data processing tasks independently, enabling real-time data aggregation across multiple financial products while maintaining manageable system complexity through modular architecture.

The patent implements an event-driven architecture where all financial product services operate at the same data processing level, publishing events to a common event stream. This equipotential design allows any financial product data to be aggregated in real-time without hierarchical bottlenecks, eliminating the month-long delays of traditional batch processing while keeping the system architecture relatively simple through standardized communication protocols.

The patent creates universal event stream compute devices and position determination compute devices that can process and aggregate data from any financial product type (deposit accounts, investment accounts, lending products, etc.). These multi-functional devices use standardized event schemas and aggregation algorithms that work across all product types, achieving comprehensive customer financial position data without proportionally increasing system complexity for each additional product type.

The patent introduces event streams as intermediary layers between various financial product services and the position determination system. Events are published to topic-based event streams that act as mediators, buffering and standardizing data from multiple sources before aggregation. This intermediary layer decouples the complexity of individual product services from the aggregation logic, allowing comprehensive data collection while maintaining manageable system complexity through standardized event processing.

The patent implements preliminary computation by maintaining running aggregations of customer financial position data in real-time through position determination compute devices. Instead of performing complex calculations only when a credit decision is needed, the system continuously pre-computes and updates customer positions as events occur. This preliminary action enables instant credit decisions without requiring intensive computational power at the moment of decision-making.

The patent uses incremental updating of customer financial positions rather than complete recalculation. When new events arrive, the system performs partial updates to existing position data, adding or subtracting event values from current positions. This partial action approach achieves real-time position determination with minimal additional computational power compared to full recalculation, enabling fast credit decisions while controlling resource consumption.