tackling big data challenge suresh karanam avp …© tech mahindra & mahindra satyam 2012 4...

15
© Tech Mahindra & Mahindra Satyam 2012 Tackling Big Data Challenge Suresh Karanam AVP & Head Global BI Consulting

Upload: others

Post on 26-Jun-2020

1 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Tackling Big Data Challenge Suresh Karanam AVP …© Tech Mahindra & Mahindra Satyam 2012 4 Defining with 3 Vs Volume (TB/sec, PB) Variety (structure) Velocity (near real-time, streams)

© Tech Mahindra & Mahindra Satyam 2012

Tackling Big Data Challenge

Suresh Karanam

AVP & Head – Global BI Consulting

Page 2: Tackling Big Data Challenge Suresh Karanam AVP …© Tech Mahindra & Mahindra Satyam 2012 4 Defining with 3 Vs Volume (TB/sec, PB) Variety (structure) Velocity (near real-time, streams)

2© Tech Mahindra & Mahindra Satyam 2012

Big Data

Challenges

Benefits

Deriving Insights

Framework

Key Steps to tackle Big Data challenges

Page 3: Tackling Big Data Challenge Suresh Karanam AVP …© Tech Mahindra & Mahindra Satyam 2012 4 Defining with 3 Vs Volume (TB/sec, PB) Variety (structure) Velocity (near real-time, streams)

3© Tech Mahindra & Mahindra Satyam 2012

Challenges

4. Enterprise

3. Government

2. Regulators

1. Citizens

Source: McKinsey - Big data: The next frontier for innovation, competition, and productivity

Page 4: Tackling Big Data Challenge Suresh Karanam AVP …© Tech Mahindra & Mahindra Satyam 2012 4 Defining with 3 Vs Volume (TB/sec, PB) Variety (structure) Velocity (near real-time, streams)

4© Tech Mahindra & Mahindra Satyam 2012

Defining with 3 Vs

Volume(TB/sec, PB)

Variety(structure)

Velocity(near real-time,

streams)

• Large volumes of data that an RDBMS

can’t handle

• Data Explosion

Structured, unstructured, audio, video, files,

click streams, web logs

Streaming data – e.g. machine generated

data

Add to all this, Complexity of Data Relationships

Page 5: Tackling Big Data Challenge Suresh Karanam AVP …© Tech Mahindra & Mahindra Satyam 2012 4 Defining with 3 Vs Volume (TB/sec, PB) Variety (structure) Velocity (near real-time, streams)

5© Tech Mahindra & Mahindra Satyam 2012

Big Data Analytics CAN make our lives better – individuals &

corporates

• Big Ones could have been better controlled by Regulatory Agencies with Big

Data Analytics – Operational Intelligence coupled with Compliance

• Enterprises benefit by having External Intelligence & Behavioral Analysis

– In turn, end consumers benefit

Operational

IntelligenceStrategic Insights

Customer

Intelligence

Page 6: Tackling Big Data Challenge Suresh Karanam AVP …© Tech Mahindra & Mahindra Satyam 2012 4 Defining with 3 Vs Volume (TB/sec, PB) Variety (structure) Velocity (near real-time, streams)

6© Tech Mahindra & Mahindra Satyam 2012

Scale of Industry Benefits from Big Data

Page 7: Tackling Big Data Challenge Suresh Karanam AVP …© Tech Mahindra & Mahindra Satyam 2012 4 Defining with 3 Vs Volume (TB/sec, PB) Variety (structure) Velocity (near real-time, streams)

7© Tech Mahindra & Mahindra Satyam 2012

Extract Intent, Life Events, Micro Segmentation Attributes

Jo Jobs

Tina Mu

Tom Sit

Chloe

Monetizable Intent

Page 8: Tackling Big Data Challenge Suresh Karanam AVP …© Tech Mahindra & Mahindra Satyam 2012 4 Defining with 3 Vs Volume (TB/sec, PB) Variety (structure) Velocity (near real-time, streams)

8© Tech Mahindra & Mahindra Satyam 2012

Extract Intent, Life Events, Micro Segmentation Attributes

Jo Jobs

Tina Mu

Tom Sit

ChloeName, Birthday, Family

Not Relevant - Noise

Not Relevant - Noise

Monetizable Intent

Monetizable IntentRelocation

Location Wishful Thinking

SPAMbots

Page 9: Tackling Big Data Challenge Suresh Karanam AVP …© Tech Mahindra & Mahindra Satyam 2012 4 Defining with 3 Vs Volume (TB/sec, PB) Variety (structure) Velocity (near real-time, streams)

9© Tech Mahindra & Mahindra Satyam 2012

Big Data Framework in Enterprise - Indicative

Data Visualization &

Analytics

Big Data Sources

Non-Traditional

Data Sources-----------------------

Social Media

Click Stream data

Account data

Web analytics

Web logs / Log files

RFID

Sensors

Machine generated

RFID

XML files

Document clusters

GPS Data

Network feeds

Cluster of

servers

XML

CSV

EDI

LOG

SQL

Text

JSON

Objects

Binary

Audio

files

Video

files

RDBMS

DW DBMS /

Analytic

Appliances

Hbase, HDFS &

MapReduce Process

Other Big

Data

Platforms

Real-Time

Traditional

Data sources

----------------------RDBMS

ERP / CRM

NoSQL DB

SaaS

Streaming……………………

1 1

2

33

3

Page 10: Tackling Big Data Challenge Suresh Karanam AVP …© Tech Mahindra & Mahindra Satyam 2012 4 Defining with 3 Vs Volume (TB/sec, PB) Variety (structure) Velocity (near real-time, streams)

10© Tech Mahindra & Mahindra Satyam 2012

Enterprise Hadoop Solution Providers

Source:The Forrester Wave™: Enterprise Hadoop Solutions, Q1 2012

Page 11: Tackling Big Data Challenge Suresh Karanam AVP …© Tech Mahindra & Mahindra Satyam 2012 4 Defining with 3 Vs Volume (TB/sec, PB) Variety (structure) Velocity (near real-time, streams)

11© Tech Mahindra & Mahindra Satyam 2012

Big Data challenges

Page 12: Tackling Big Data Challenge Suresh Karanam AVP …© Tech Mahindra & Mahindra Satyam 2012 4 Defining with 3 Vs Volume (TB/sec, PB) Variety (structure) Velocity (near real-time, streams)

12© Tech Mahindra & Mahindra Satyam 2012

Big Data challenges – Cont’d…

Source: TDWI Best Practices Report – Fourth Quarter 2011

Page 13: Tackling Big Data Challenge Suresh Karanam AVP …© Tech Mahindra & Mahindra Satyam 2012 4 Defining with 3 Vs Volume (TB/sec, PB) Variety (structure) Velocity (near real-time, streams)

13© Tech Mahindra & Mahindra Satyam 2012

Key steps to tackle Big Data challenges

Cross-skill related skilled-personnel to gain experience in Enterprises & IT Organizations

Think BIG, Start small with most relevant use cases by demonstrating Proof of Value;

improvise on the processing power of Big Data solutions with the right architecture

Gain end users’ confidence & thereby business sponsorship for a scalable Big Data

Platform

Preparation of data, developing concepts, deployment of right models & data stewardship

Beware of the obvious presence of false information due to manual interventions

1

2

3

4

5

Converse with fellow Big Data consultants & share experiences6

Page 14: Tackling Big Data Challenge Suresh Karanam AVP …© Tech Mahindra & Mahindra Satyam 2012 4 Defining with 3 Vs Volume (TB/sec, PB) Variety (structure) Velocity (near real-time, streams)

14© Tech Mahindra & Mahindra Satyam 2012

Benefits Business & Technocrats equally…

Source: McKinsey - Big data: The next frontier for innovation, competition, and productivity

Page 15: Tackling Big Data Challenge Suresh Karanam AVP …© Tech Mahindra & Mahindra Satyam 2012 4 Defining with 3 Vs Volume (TB/sec, PB) Variety (structure) Velocity (near real-time, streams)

15© Tech Mahindra & Mahindra Satyam 2012

Thank YouContact: [email protected]