Realise the promise of big data. - SAS...Page 3 © Hortonworks Inc. 2011 –2015. All Rights...
Transcript of Realise the promise of big data. - SAS...Page 3 © Hortonworks Inc. 2011 –2015. All Rights...
Realise the promise of big data.
Simon [email protected]
© Hortonworks Inc. 2011 – 2015. All Rights Reserved
Page 2 © Hortonworks Inc. 2011 – 2015. All Rights ReservedPage 2 © Hortonworks Inc. 2011 – 2015. All Rights Reserved
Hadoop core concepts.Purpose and outcome.
Store an unlimited quantity and variety of data in a
single place, regardless of it’s format or origin.
Process - merge, interact, sort, refine, normalise,
index, expose and ANALYSE that data.
Operationalise that data for your benefit.
Page 3 © Hortonworks Inc. 2011 – 2015. All Rights ReservedPage 3 © Hortonworks Inc. 2011 – 2015. All Rights Reserved
Hadoop core concepts:Part of the modern data architecture.
AppApp App App
App
App
Hadoop
CLICKSTREAM SENSOR SOCIAL MOBILE GEOLOCATION SERVER LOG
Batch Interactive Search Streaming Machine Learning
EXISTING
Page 4 © Hortonworks Inc. 2011 – 2015. All Rights ReservedPage 4 © Hortonworks Inc. 2011 – 2015. All Rights Reserved
Why Hadoop & SAS?
Challenges & Opportunities.
Mainframe
Data
Device
Data
Machine
Data
Product
Design
Social
Mapping
Factory
Yields
Defect
Detection
Human
Data
Archive
Data
Proactive
Repair
Disaster
Mitigation
Investment
Planning
Next
Product
Rec’s
Store
Design
Risk
Modeling
Ad
Placement
Inventory
Predictions
Sentiment
Analysis
Ad
Placement
Basket
AnalysisSegments
Customer
Support
Supply
Chain
Cross-
Sell
Customer
Retention
Vendor
Scorecards
Optimize
Inventories
Page 5 © Hortonworks Inc. 2011 – 2015. All Rights Reserved
Audience understanding:Find – Understand – Engage - Measure
Rogers’ Media Audience Platform: Integration and analysis of all data collected across the organization
Query all data in one location Blend of online and offline data, subscription, ecommerce, loyalty programs, etc.
Land massive click stream log files, 100+ M records / day, 30 million unique IDs / month
Use 100% of the data for Analysis and Visualization instead of smaller random samples (over sampling)
Identified and modeled more than 600 relevant web characteristics out of a field of 75,000 with SAS
Deliver products and solutions with relevance and timeliness.
Why Hadoop
and SAS?
Single View of
the customer.
TelcoRogers Media is a
subsidiary of Rogers
Communications,
which owns Canada's
largest publishing
company.
Page 6 © Hortonworks Inc. 2011 – 2015. All Rights Reserved
Operationalise manual processes at scale:Difficulty identifying coding errors among 300K daily claims
Using Analytics and Hadoop to improve reimbursement revenue and health outcomes
HDP + SAS: Marrying and analyzing numerous pool of data stored in multiple silo’s—including gross margins, taxes, customer claims and policy premiums—to determine the company's potential exposure.
Ability to crunch several terabytes of data, and then revise, recalculate and report on that data on a regular basis.
Removing costly manual intervention and increasing accuracy.
Insurance
Healthcare
Large US medical
insurer
Why Hadoop
and SAS?
Data Discovery
Page 7 © Hortonworks Inc. 2011 – 2015. All Rights Reserved
Unable to monetize data.Fragmented data silo’s.
Data flows into one data lake with centralized
security policies, reducing storage costs
Multiple data sets extracted from source platforms with single
point of security & privacy for de-identification, masking,
encryption, authentication and access control.
All departments have access to the same cross-sell data
Seamless integration with SAS
Banking
One of the largest US
banks
Why Hadoop
and SAS?
Data Discovery
Page 8 © Hortonworks Inc. 2011 – 2015. All Rights ReservedPage 8 © Hortonworks Inc. 2011 – 2015. All Rights Reserved
Hadoop & SAS:
Why we do what we do…
Hortonworks HDP is Apache Hadoop.
SAS is the leader in analytics.
Hadoop on YARN – A Data Operating System
CLICKSTREAM SENSOR SOCIAL MOBILE GEOLOCATION SERVER LOG
Batch Interactive Search Streaming Machine Learning
EXISTING
SAS Data Management & Analytics Suite
Skills & Processes.
Use existing skills and processes against
net new data.
Data Management.
Ingest, transform, cleanse and tag data
within the Hadoop eco-system.
Operationalise Analytics.
Apply analytics and rules to pinpoint event
relevance and urgency with continuous
pattern detection.
Page 9 © Hortonworks Inc. 2011 – 2015. All Rights ReservedPage 9 © Hortonworks Inc. 2011 – 2015. All Rights Reserved
Profile CleanseIn-Memory
AnalyticsScore
Ingest
SAS® High-Performance
Analytics
SAS® Visual Statistics
SAS® In-Memory Statistics
SAS/ACCESS®
interface to Hadoop
SAS® Data Loader for Hadoop
Data
MartsPredictive
Modeling
Model
Scoring
Data
Preparation
Transform SAS® Scoring Accelerator
Stream
SAS® Event Stream
Processing
Hortonworks & SAS combined…Open Source Platform – Highly functional software solutions.
Page 10 © Hortonworks Inc. 2011 – 2015. All Rights Reserved
708 Active
Subscription
Customers
The Hortonworks Methodology.Open Source model defined
Develop
Distribute
Work with the open source
community, develop Hadoop projects
to meet business requirements
Distribute tested, packaged
versions of Hadoop to
end customers and their users
Support
Architect
Introduce business
requirements for planning
future iterations of Hadoop
Provide end user support
to deployed Hadoop clusters,
capturing new requirements
Page 11 © Hortonworks Inc. 2011 – 2015. All Rights ReservedPage 11 © Hortonworks Inc. 2011 – 2015. All Rights Reserved
Summary
Why Hortonworks and SAS?
We provide our customers analytics against 100 %
of the relevant data.
Provide customers a comprehensive insight into any
entity
Provide customers the ability to make data derived
decisions both pre & post transactional.
Page 12 © Hortonworks Inc. 2011 – 2015. All Rights Reserved
Questions?