Building a Real-Time Complex Event Processing Platform for Investment eCommerce

Defined and delivered a distributed Lambda Architecture integrating Spark and Cassandra to enable real-time content recommendations, behavioural alerts, and analytics within a global securities trading platform.

The Client

A top-tier global investment bank and financial services institution, recognised as one of the largest wealth and asset managers in the world. The organisation manages approximately US $7 trillion in assets under management, serving institutional, corporate, and private clients across all major financial markets.

With a workforce of 100,000+ professionals and operations spanning 50+ countries, the bank maintains strategic hubs in key global financial centres, including a major presence in the heart of the City of London. The institution is classified as a global systemically important bank (G-SIB), reflecting its scale, complexity, and critical role in international capital markets.

$7T+

Assets under management

100K+

Employees worldwide

50+

Countries of presence

The Challenge

The bank's global multi-asset digital platform unified research, analytics, pricing, and execution into a single interface used by clients, sales teams, and traders worldwide. While this environment generated significant volumes of behavioural, content, and market events, it lacked a dedicated capability to analyse and react to them in real time. Business stakeholders identified four priority needs:

Deliver personalised content recommendations based on user search behaviour.
Provide sales teams with contextual alerts derived from client activity.
Detect barrier price knock-in/knock-out events and notify relevant users.
Generate structured analytics on research readership and engagement.

Addressing these requirements required introducing a scalable Complex Event Processing (CEP) layer capable of correlating real-time and batch events across distributed systems, and integrating seamlessly with the existing global alerting framework.

Our Solution

Designed and implemented a distributed Lambda Architecture introducing Complex Event Processing (CEP) capabilities into the bank's global securities platform, enabling real-time and batch event correlation at scale.

Event-Driven Architecture

Defined and deployed a scalable CEP framework leveraging Spark Streaming and Cassandra to process high-frequency behavioural and market events. The architecture supported both real-time streams and batch ingestion and processing.

Personalised content engine

Implemented streaming pipelines to analyse user search activity and interaction patterns, dynamically generating tailored content alerts and contextual recommendations.

Market Event Detection

Integrated end-of-day pricing feeds to detect barrier price knock-in/knock-out conditions, automatically triggering notifications to traders and relevant sales users.

Research Usage Analytics

Built batch analytics processes to measure readership patterns, aggregating author- and publication-level engagement metrics to support editorial and commercial insights.

Enterprise Integration & APIs

Integrated the CEP layer with existing ESB, monitoring, and REST services, ensuring seamless interoperability with the bank's alerting infrastructure and operational systems.

DevOps & Industrialisation

Embedded Spark within the bank's Continuous Deployment framework, enabling iterative Agile delivery, production rollout across 7 global data centres, and knowledge transfer to internal platform teams.

Secure Kappa architecture for investment event processing

The Value

The introduction of a production-grade Complex Event Processing capability transformed the platform into a real-time, behaviour-driven digital ecosystem, enhancing content relevance, improving responsiveness to trading conditions, and establishing a scalable foundation for future real-time analytics innovation.

CEP Use Cases

Four real-time and batch use cases industrialised and deployed in production.

Global Data Centers

Distributed CEP architecture deployed across seven international data centres.

Project duration

<6mo

End-to-end CEP platform delivered from design to production in under six months.

Real-Time User Engagement

Enabled dynamic, personalised content recommendations and contextual alerts driven by real-time analysis of user search and interaction behaviour.

Faster Market Responsiveness

Enabled timely detection of key market events and automated notifications, improving reaction speed within trading and sales workflows.

Scalable Event-Driven Foundation

Established a production-grade real-time analytics layer, creating a reusable foundation for future streaming, personalisation, and advanced analytics initiatives.

View More Success Stories