Using Event Driven Architecture to Transform Core Banking...Reactive API Gateway Transaction Svc....

Confidential and Proprietary. Copyright © Accenture, LLP. All rights reserved.

Using Event Driven Architecture to Transform Core Banking


Acceleration of digitization:• Traditional companies are becoming software companies,

whether they like it or not!• Customer experience demands are shifting• More interaction with machines• Machines are making simple decisions (and are better at it)

Banking is quickly shifting to meet the challenge. The question is, are they quick enough?


Key Premises

Regulation is here to stay; “you can’t fight the man”

Finance emits hundreds of events, many worth acting on!

We can’t quickly replace legacy with greenfield

Customers demand ever better user experience at an increasing pace

Confidential and Proprietary. Copyright © Accenture, LLP. All rights reserved. Confidential and Proprietary. Copyright © Accenture, LLP. All rights reserved.

Case Study


The Business Situation of nearly every bankRising Mainframe CostsRapidly increased over last years. Ongoing increase of + 5% CPU usage per year

2013 2015 2015NochangeScenario

2015Goal

PlannedCostReduction

+5%CPUusageperyear

More Devices more Channels80% of online transactions from mobile devices

~60% of overall transactions are read-only request

Slow InnovationNew products and services take months/years to deliverThe business watches fintech with increasing fright

Banks are becoming software companies whether they like it or not, but often their digital product leaves a lot to be desired

?!


Hadoop Cluster

High Level Architecture

Mainframe

Application(s)

DB2Agent

z/OSAgent

Linux

MasterSecondary

NodeNode

NodeNode

Node

Loadbalancer

Writ

e-Pa

th(C

opyb

ooks

…) Read-Path

(REST/JSON)

NewExisting

Microservice

Even

t Hub

Fraud Detection

MasterPrimary

Microservice


Cost Savings + Market Leading CapabilitiesOnline Transactions

Red

uced

O

nlin

e Tr

ansa

ctio

ns

Mainframe CPU

Red

uced

M

ainf

ram

e C

PU

NewNew

• 50% reduction in mainframe CPU, 50% reduction in ongoing mainframe cost

• went from planned expansion to reduction in hardware

Event Hub

Fraud Detection

Trading

API API APIAPI Gateway

Billing Others

Microservices

Systems of Record

Channels

Event StreamAPI API API

SCI Digital Marketplace

• Implemented fraud detection capability in 4 weeks, well short of 7 month deadline J

• went from planned expansion to reduction in hardware


A Reference Architecture


Logical Components

Reactive API Gateway

Transaction Svc. User Svc. New ServiceProcess Svc.

Transactions User / Prefs. Forms

Legacy Services

Transactions

New Store

CD

C

Rep

licat

ion

Event Hub

Inte

rnal

Exte

rnal

User/Mobile User/Web Customer Support 3rd Party Services

Synchronous REST Calls

Streaming Messages

Apple/Google Push API

Notification Service

Other Utility Svcs.

Fraud Detection (Analytics)


Technologies

Reactive API Gateway

Microservice Microservice Microservice

Transactions User / Prefs. Forms

Legacy Services

Transactions

CD

C

Rep

licat

ion

Inte

rnal

Exte

rnal

User/Mobile User/Web Customer Support 3rd Party Services

Synchronous REST Calls

Streaming Messages

Apple/Google Push API

Notification Service

Other Utility Svcs.

Fraud Detection (Analytics)


Inside a Microservice


Extending to Other Industries: Decouple, Isolate, Hollow, Replace

Using Event Driven Architecture to Transform Core Banking...Reactive API Gateway Transaction Svc....

Documents

Transcript of Using Event Driven Architecture to Transform Core Banking...Reactive API Gateway Transaction Svc....