cisco data virtualization & connected analytics · service provider. 10 levels application 6...

20
1 Cisco Data Virtualization & Connected Analytics Ché Wijesinghe, October 10 th , 2014

Upload: others

Post on 16-Apr-2020

15 views

Category:

Documents


0 download

TRANSCRIPT

1

Cisco Data Virtualization & Connected AnalyticsChé Wijesinghe, October 10th, 2014

2

IoE Changes Everything…

3

Internet of Things

Internet of Things

Location of Data

Location of Data

Velocity of Data

Velocity of Data

Distributed Data

Distributed Data

StreamingData

StreamingData

DataData

Big DataBig Data

Worldwide Data Volume is Doubling Every 2 Years

CloudCloud

InfrastructureInfrastructure

ApplicationsApplications

4

Busin

ess L

evera

ge o

f D

ata

Busin

ess L

evera

ge o

f D

ata

DataData

� Agility

� Cost Reduction

� Competitive Advantage

� Agility

� Cost Reduction

� Competitive Advantage

ObjectivesObjectives

25% of big data implementations will fail to deliver business value resulting from performance problems due to inadequate network infrastructure.

6

• Cisco has deep experience analyzing varieties of Big Data from its own global operations to improve business and customer outcomes

– Contact center analytics – Cisco TAC

– Collaboration analytics on Cisco’s own operations

– Data warehouse offload

• Cisco technologies drive customer Big Data analytics– Cisco data analytics platform – easily consumed analytics and

visualizations

– Cisco Data Virtualization – unified view of diverse data for analytics and applications

– Cisco Prime Real-Time Analytics – streaming and IoT applications

7

Channel Profitability Analysis

Customer Profitability Analysis

360 Degree View of Customer

Marketing Campaign Analysis

Sales Force Automation

System (Legacy)

Campaign Management System

Sales Force

Automation System

(Cloud)

Trades and Asset

Management System

(Cloud)

Sales Force

Automation System

(Cloud)

Trades and Asset

Management System

(Cloud)

Investment Account Master

Website Clickstream Data Store

Email Management System

� Better business insights� Faster response times� Cost savings through streamlined approach

� Better business insights� Faster response times� Cost savings through streamlined approach

Data VirtualizationData Virtualization

8

Unified View of Data

9

Packaged, network-enriched analytics leverage Cisco network technologies and data for targeted verticals

Retail

Store operations and customer service

Collaboration

Organizational effectiveness, collaborative selling

Contact Center

Personalized contact center service in real-time

Network

Optimized networks for service agility

Cisco Connected Analytics Solutions

Customer stickiness and operational efficiency

Service Provider

10

Levels

Application(Reporting, Analytics, Control)6

Data Abstraction(Aggregation & Access)5

Data Accumulation(Storage)4

Data Element (Analysis & Transformation)3

Connectivity(Comm & Processing Units)2

Physical Devices(The “Things” in IoT)1

Data at Rest

Data in Motion

Sensors, Devices, Machines,Intelligent Edge Nodes of all types

TargetedSolutions

Center

Edge

UCS, Switches,Routers

Data in Motion

Data Virtualization

VerticalIndustry

Solutions

PrimeAnalytics,Whiptail

11

Data Abstraction

(Aggregation & Access)5Cisco Data Virtualization

1. Provide views of data in the manner that applications want

2. Combine data from multiple sources, simplifying the application

3. Filtering, selecting, projecting, and reformatting the data on behalf of the app

4. Reconciling differences in data shape, format, semantics, access protocol, and security

CIS

UCS

Abstracting the data interface for applications

12

Data Abstraction

(Aggregation & Access)5

Malaysia Alaska

WAN

Houston

GeologistData

Analyst

13

� Improve overall risk management operations

� Improve information and report quality, improve data accuracy

Objective

� Data scattered with varying levels of quality

� No consistent method of providing an aggregated view of financial activities

Challenges

� Pull the data once and reuse it many times

� Eliminate enterprise applications going directly to the data sources

� Create a mapping between the virtual schema and the logical model that is used for data governance

� Provide a centralized repository for data quality

� Map from the virtual to the physical schema’s for audit and data lineage

Solution: Cisco Data Virtualization Suite

� Improve employee productivity and reduce costs

� Improve information and report quality

� Reduce the number of data audits

� Reduce the number of data loaders

� Reduce the operating costs by facilitating the consolidation of systems

Results Cisco Information ServerCisco Information Server

Credit Risk Market Risk Trade Compliance

BORTSystems

ReferenceData

PurchasedData

LOB Data

14

� Revenue growth: the business wanted to increase the effectiveness of upsell programs

� Looking for trends in customer behavior to drive marketing insights for email campaigns

Objective

� Want to analyze integrated collection of clickstream data, system usage data, and dimensional data for PlayStation 3

� Analysis frequency is extremely limited

� Calculations can only be run once/month or once/week; need this data daily

� Combine different data sources and make it look like unified source; limit replication

Challenges

� A flexible data model/abstraction layer across all source systems

� Bridge dimensional data from Exadata with clickstream data and system usage data from Hadoop

� Enhanced agility through a flexible data delivery infrastructure

Solution: Cisco Data Virtualization Suite

� $9M revenue increase due to business value acceleration

� IT staff savings of $415,000 on first project

� Infrastructure cost avoidance of $304,000 on first project

� 7.2x ROI in first year

Results Cisco Information ServerCisco Information Server

EcommerceAnalysis

System Usage Analysis

SegmentationAnalysis

Exadata SQL Server Oracle

15

� The IT Problem: Spend on storage growing far faster than budget allows

Objective

� Business Data User Problem: “I don’t care where the data is coming from, or where it is stored, but I do care when I see critical project estimates and durations exceeding my ability to pay or to wait on results!”

� There are 2 major enterprise data warehouses: Wireless and Wireline; 20 additional DW’s; total = 20 PBs

� Source DW cost = $100K/TB; destination data lake = $2K/TB; Migrating a typical set of tables from source to target might take an average of 10 months.

Challenges

� With Cisco Information Server, an abstraction layer can be built and consuming connections much more easily made to this layer, requiring half the time.

� One Example Project: one set of tables is growing at 10TB/mo; beginning after 10 months the savings will be $980K/month

Solution: Cisco Data Virtualization Suite

� 4-5X times faster to build a virtual solution

� 5 months early migration yields $4.9M for just this one migration project.

Results

Director of Data Strategy and Support

“When the business have a new data source or federation project, we can now deliver results in 1/5th the time. Even if we eventually take the data to the warehouse, total data

integration project costs are still 50% of the old direct-to-warehouse solution.”

Director of Data Strategy and Support

Cisco Information ServerCisco Information Server

Billing Analysis Ordering Analysis Customer Care Analysis

Wireline Data Wireless Data Customer Data

16

Initiative

◦ Bank-wide Data Virtualization

Challenges

◦ Up-to-the-minute information needed for multiple risk management, trade order management, and debt/equity market research applications

Solution

◦ Bank-wide Data Virtualization Layer via a services-oriented-architecture using Composite accesses, abstracts and delivers required data

◦ 25 applications, 200+ Sources

Results

◦ 250% ROI in 3 months elapsed time

◦ 2% revenue increase within Corporate Investment Bank

◦ 50-60% reduction in integration design and development time for new applications and portals

◦ 25% increase in object reuse for downstream BI reporting projects

“Typically, data integration is 50% of the elapsed time of each of our large portal and applications projects. We are able to cut the data integration development time by 50-60%

using the Composite Information Server.”

VP, Corporate Investment Banking Technology

Product Data(Sybase)

Coverage(SQL Server)

Holdings &Offerings

ReferenceData

(Teradata)

Trade OrderManagement

RiskManagement

Debt / EquityMarket Research

Cisco Information ServerCisco Information Server

17

� Improve the speed and quality of risk and compliance analytics.

� Access and combine a vast variety & volume of data, including semi-structured and structured data, and make historical data queryable.

Objective

� Residential mortgages are very sophisticated products in the financial industry, with many complex interactions.

� Extended lifecycle of a mortgage contributes to complexity; each step in the process involves multiple systems, parties, and relationships.

Challenges

� Single provisioning hub, presenting data view from multiple underlying data stores.

� Provide snapshots of detailed loan data for response to audits, inquiries and disputes.

� Provide access to complete data without need for traditional suite of tools.

� Provide accurate and consistent information to Regulators, avoiding penalties.

Solution: Cisco Data Virtualization Suite

� Respond swiftly to business events.

� Reduced number of provisioning points from greater than 100 to less than 20.

� Reduced cost to acquire, provision and consume data.

� Moved from 70% time spent in data acquisition to 70% data usage and delivery.

� 60% reduction in infrastructure costs.

Results

Cisco Information ServerCisco Information Server

Loan Servicing Operational Analysis

Regulatory Reporting

Reference/ Master Data

Risk Finance Data TransactionsMortgage Data

18

Traditional Analytics Days, Hours

Single Data WarehouseTraditional

Data WarehouseTraditional

Data Warehouse

Structured DataStructured Data

Big Data Store

Big Data Store

Fast DataFast Data

Unstructured Data

Unstructured Data

Big Data AnalyticsHours, Minutes

Multiple Data Sources

Next Generation AnalyticsSeconds, Milliseconds

Data Everywhere

RFID

Real-time,

StreamingIntelligent Network

DVDV

19

HCS

MicrosoftSuite aaS

DRaaS

PaaS

IaaS

EnterprisePrivateClouds

Public Clouds

Partner CloudsCloud Services and

ApplicationsInterCloud Fabric

APIs

Portal

APIs

APIsOpenStack

Meraki

Security

Analytics

vDesktop aaS

WebEx

HANA aaS

IOE aaS

Collaborationand Video

Big Dataand Analytics

Native CloudApplications

EnterpriseWorkloads

20

TOMORROW starts here.