tcs point of view session - analyze by dr. gautam shroff, vp and chief scientist, tcs
TRANSCRIPT
Drive to take more actions/decisions in
presence of more information and less time
Evolution of Enterprise Analytics
Isolation:What has
happened?
In-time:What is
happening?
Integration: Why has
it happened?
Intelligence:What will happen?
What are others saying?
Information:Harnessing
Knowledge
“any fool can know … the point is to understand.”- Albert Einstein and … the goal of understanding is to predict
Reactive Intelligence Predictive Intelligence
*courtesy Pawan Sinha, MIT
“any fool can know … the point is to understand.”- Albert Einstein and … the goal of understanding is to predict
Reactive Intelligence Predictive Intelligence
*courtesy Pawan Sinha, MIT
“any fool can know … the point is to understand.”- Albert Einstein and … the goal of understanding is to predict
Reactive Intelligence Predictive Intelligence
*courtesy Pawan Sinha, MIT
sampling P(X) manually => infinite time / infinite # people!
m attributes, each with d possible values: O(d2m) ‘cubes’
for m=40, d=10 this becomes 1080 > # atoms in the universe
so – BI folks need to learn analytics
Customers( x1… xm)
Big Data is about ‘wide’ data
Random-access to data is poor, even in memory!
Map-reduce based procedures exploit this.
network
speed
distributed
processing works
in-memory DB
no panacea
POV 1 : Big Data is here to stay and will be an increasingly significant arena of competitive
differentiation
POV 2 : There are two fundamental aspects to Big Data : Harnessing: The Technology required to
Manage Big Data and Harvesting : The Technology required to analyze and derive insight from Big
Data.
POV 3 : Big Data Technology Platform can solve traditional Data Problems as well and is not
limited by the use of Big Data itself.
POV 4 : The current innovation landscape is vast, varied with multiple products and offerings. We
can expect natural Consolidation in next 2-5 years.
POV 5 : Unstructured Data cannot be consumed in its raw form. It is essential to convert it
into consumable structured form for useful interpretation
POV 6 : Fusion of Unstructured and Structured Information is creating the need for a new science
stream: Data Science which requires both Business context and Hard Science
POV 7 : Big data is in the incubation Phase for most of the organizations. Only the likes of
Google, Yahoo, Amazon, Facebook are matured adopters.
POV 8 : Enterprises will have to undergo business process adjustments / redefinition both for
upstream and downstream connect (consumption) on big data, i.e. harnessing and harvesting
Event Detection Engine
Data Harmonisation
Causal Analytics
Framework
Topic Evolution in News
Email Mining …
Sensor Pattern Matching
Sensor Data Motif Discovery
Temporal Event/Sensor Rules
Geo-Spatio-Temporal
Patterns/Motifs/Rules
Sensor-stream Databases
Collective Entity Resolution
Relationship Discovery
Searching Linked-open-data
Federated Object Discovery
TCS Big Data
AcceleratorsSentiment Analysis,
Social Media Adaptors,
Data Connectors,
Video Analytics, Utiliti
es
TCS Active Archival
Archival using
Hadoopstorage.
Abundant space. Warm
data
TCS Meta Data
Manager
Searchable platform to manage the metadata of Hadoop data
across clusters
TCS DataMigration
Tool
Fast, secure data
movement in/out of
Hadoop from any source (m/f, oracle
etc.)
TCS Sensor Data
Analytics
Receive, store and analyze any type of sensor / log
data
TCS Perigon™
Provide a confluence of customer data and analytics
using enterprise as well as social
data (Customer 360
view)