Big Data - A Real Life Revolution

#tduniv
Big Data
A Real Life Revolution
April 2014, Teradata Universe
Manuel Sevilla
Global CTO
Business Information Management
@msevillatweets

2
Big Data & Analytics
Copyright © 2014 Capgemini. All rights reserved.
What is Big Data?
76 million
smart meters
in 2009…
200m by
2014
2+ billion
people on the
Web by end
2011
100s of
millions of
GPS enabled
devices sold
annually
4.6 billion
camera
phones world
wide
30 billion
RFID tags
today (1.3B in
2005)
12+ TBs
of tweet data
every day
25+ TBs
of log data
every day
Many PBs
of data every
day
80%
Of world‟s data
is unstructured
The 3Vs: Volume, Variety & Velocity

3
The real 3V explanation
 Volume: BIG DATA
• More and more devices and
• Each device generates more and more data
 Variety: ALL DATA
• Non structured and structured data together
• This is about an interconnected world with many external
partners working with no or low-modeled data
 Velocity: FAST DATA
• This is about information and insights value
• Data and insight value decreases every minute!
1 0
0 1 01 0
0 1
0 1
1 1 0
1 0 1 1
0 0 1 1 1 0 1
0 1 0 0 1 1 0 1
1 0 1 0 1 1 0 0
1 0 1 1

4
In-memory has changed the game
An in-memory appliance
40 x86 cores, 1TB of RAM
For only 100 K EUR !
Moving a standard application from in-disk to
in-memory (without redesign) means a performance
increase between 100 and 1000 times faster
1 to 100 ratio: 2 minutes become 1 second
1 to 1000 means that
48 hours process may run in 3 minutes !

5
5760 cores for $3000 !!!
With two GK110 chips,
TITAN Z is powered by
a total of 5,760
processing cores, or
2,880 cores per GPU
Analytical
power is not
a limit
anymore !

6
From Predictive Asset Management…
Dust
from
dirty
wagons
Dynamic forces from trains
rubbing the ballast particles
together
Ballast choked
with fines – poor
drainage
The stiffness of
the underlying
formation
Voids under track
Embankments can degrade
through drying out and
shrinkage esp. in SE
Environment
dead vegetation
Strength of
underlying earthwork
Wet bed

7
to Predictive Human Healthcare

8
BI Appliances Hadoop
 Expensive dedicated HW
 Built for performance
 Designed for high volumes (e.g. 10s of TB)
 High availability (limited due to price pressure)
 Initially developed using Relational Data bases
 Very mature solutions (skills, SW, HW, administration)
 Designed for modelled and structured data
 Business As Usual ways to design, build and deliver
 Teradata, Exadata, Netezza, HANA...
 Commodity PCs
 Built for extreme scalability (Batch oriented)
 Designed for extreme volumes (10s of PB and more)
 Very high availability
 Initially developed for web ranking
 Not as mature
 Hadoop = Data is distributed over many machines
 MapReduce = Computing is distributed and executed
where data is (grid solution)

9
Treatment Key
Object
SQL Database Search In
memory
Hadoop is first of all an extremely sustainable solution
Hadoop Distributed File System
Impala
Spark
HBase
Map
Reduce HAWQ
SolR
SAS
HPA
HIVE, PIG, JAQL,
R support, ETL…
H
Catalog
Stinger
Connectors
Flume Storm Sqoop

10
IT World was split in 2 categories
Historical
Data
Streaming
Data
(Events)
GB
TB
PB
GB/s
MB/s
KB/s
Day Hour Min Sec SubSec
Data Warehouses
OLTP
Databases
Volume
Events Response Time

11
Big Data is about new Solutions, new opportunities
Historical
Data
Streaming
Data
(Events)
GB
TB
PB
GB/s
MB/s
KB/s
Day Hour Min Sec SubSec
Volume
Events Response Time
Data Warehouses
OLTP
Databases
Event Processing
Tools
Hadoop
In-memory
Databases

12
What is really innovative behind the Big Data buzz word ?
Hadoop & NoSQL : Low cost ways to store, manage and analyze
massive volumes of data
Cloud : Ability to rent, reduce Infra cost, reduce time to market
Event Processing tools : Ability to analyze and detect trends in real
time streaming events (monitoring, Next Best Action, Fraud...)
In-memory technologies : A new way to guarantee response time
even for very complex calculations
Explosion of Analytics usage : R is open source, High performance
statistics make them usable by every needed process
Multiple external data sources : Easier to tackle new data sources eg.
social media, traffic, GPS sensors, open data.

13
Multiple internal views – consistently compromised
Corporate
Ad-hoc
LOB
management
Operations
Market
Operations LOB mart Spreadsheets
Line of business
Transactional systems
CRMERP PLM
EDW
Fit
Detail
Freshness
Fidelity
Corporate
ODS
Web

14
The Business Data Lake Solution
HDFS
Load everything
Keep the history
Business driven
Country
Sales
Asian Marketing
campaign
Operations
data mart
Distill
Structure
Transactional systems
CRM PLMERPMeters GridWeb Social Media MarketSupplier

15
You only need to be a few %
points better than your
competitors

16
The New Analytics Organization
 A Proof of Concept (PoC) is usually delivered in months
 Fail fast methodology!
 This has to be considered as Research & Development : discovering new
patterns, new ways to reduce cost, improve customer experience, work on
Predictive Asset Maintenance, Fraud Detection
 Massive use of Cloud
 to reduce Time to Market
 CAPEX vs OPEX , as a service mode
 Very close to business and IT, often a Mixed team (internal + suppliers)
 Dedicated to deliver new insights fast
 Full access to many company data sources
 Deep knowledge of IT data sources and modeling
 Deliver temporary solutions, planed to be used as prototyping for
industrialization
Fast Analysts Teams
PoC Analysts teams

17
Data monetization : new revenue?

18
Data Privacy: 3 levels
May be federal, country or state level
May concern usage, storage, Information Lifecycle Management
May include legal requisitions (as with telco companies)
Level 1 : Law & Regulation
This is the way you want and you do not want to use all this information!
May be defined per BU, country…
May include partners / suppliers
Level 2 : Company level
Every customers wants be able to accept or reject any information usage
This may be linked to a contract, a checkbox, a call, a spam declaration…
This is not business as usual
You need to give to your customers the ability to define their own data access and
usage wishes (authorizations) !
Level 3 : Personal level

19
Our Big Data Service Center Framework
Deliverymanagement
Technical coordination
Service Center Governance
and PMO
Functional Coordination
Lean / Continuous
Improvement / Best
Practices
Functional & Technical
expertise
Factories governed by people, process and technology
POC Design Build RUN
Transformation – Big Data Business Cases
New Big Data development
Data usage (API, Analytics…)
Adding new data sources
Demand management Resource
management
Quality and
performance
management
(Progress and quality)
Configuration and
release management
Infrastructure
architecture
SandboxLink with legacy
Datacenter
(countries)
Privacy management
Business Analysis
Datalab
Industrialization GO Live
Innovation
Big Data Governance
Design authority
(Big Data architecture/roadmap)
BI vs Big Data
governance
Change management
Prioritization and
sponsorship
Transversal Delivery Strategy & Innovation
Infrastructure
Data Scientists
Analytics
Privacy & ILM
VM Package
Network
Management
Architecture
Product / Solution
choice
Delivery Model
selection
Information Lifecycle
Management
Support and
Administration
Analytics as a
Service

20
SMAC is dead. Long live SMAC!
SMAC New SMAC
Social Security
Mobility Movement
Analytics Artificial Intelligence
Cloud Creativity

21
Capgemini Global BIM Service Line
Driving innovation with its technology partners
 Capgemini „s global reach with operations in 44
countries and a focus on BIM with over 9600 BIM
practitioners.
 A uniquely integrated approach to Information
Strategy based around the Capgemini “Intelligence
Enterprise”.
 Deep Industry sector knowledge supported by Sector
Specific BIM offerings.
 Capgemini‟s best-in-class Rightshore® capability for
BIM for development and management of BIM –
4000 BIM experts in India CoE.
 A unmatched (and vendor independent) depth of
technology experience. Capgemini works with all the
major BI software vendors to deliver solutions
appropriate to the customer‟s needs.
850+ M EUR revenue in 2013
Europe:
South Africa
Argentina
Brazil
Mexico
United States
Canada
Saudi Arabia India
Australia
China
Morocco
Austria
Finalnd
France
Italy
Germany
Norway
Netherlands
Poland
Spain
Sweden
Switzerland
UK

22
Capgemini Group focuses on Big Data
Capgemini Group focuses on Big Data up to top level executive with his CEO, Paul Hermelin,
to support French Industry Ministry on Big Data project (34 French‟ Industrial projects)
www.redressement-productif.gouv.fr/files/la-nouvelle-france-industrielle.pdf#page=27
Paul Hermelin, Capgemini Group CEO, on live in
BFM Business® on 11/10/13 presenting Big Data
project for French Industry government .

23
Contact
Manuel Sevilla
Global CTO
Business Information Management
@msevillatweets
#tduniv

The information contained in this presentation is proprietary.
Copyright ©2014 Capgemini. All rights reserved.
Rightshore® is a trademark belonging to Capgemini.
www.capgemini.com/bim
About Capgemini
With more than 130,000 people in over 40 countries, Capgemini
is one of the world's foremost providers of consulting, technology
and outsourcing services. The Group reported 2013 global
revenues of EUR 10.1 billion.
Together with its clients, Capgemini creates and delivers
business and technology solutions that fit their needs and drive
the results they want. A deeply multicultural organization,
Capgemini has developed its own way of working, the
Collaborative Business Experience™, and draws on Rightshore®,
its worldwide delivery model.

Big Data - A Real Life Revolution

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Viewers also liked

Viewers also liked (13)

Similar to Big Data - A Real Life Revolution

Similar to Big Data - A Real Life Revolution (20)

More from Capgemini

More from Capgemini (20)

Recently uploaded

Recently uploaded (20)

Big Data - A Real Life Revolution