driving the digital enterprise with big data€¦ · sample case studies •customer: a consortium...
Post on 28-Jun-2020
2 Views
Preview:
TRANSCRIPT
ShareInsights
Driving the Digital Enterprise with Big Data
Mukund Deshpande – GM, ShareInsights
Confidential. Copyright © 2017 Accelerite. All rights reserved.
Big Data Market & Positioning1
What are the Key Challenges ?2
What is ShareInsights ?3
Demo4
AGENDA
© 2016 Accelerite. All Rights Reserved. Confidential. Copyright © 2017 Accelerite. All rights reserved.
Enterprise
Data Stacks
ENTERPRISE DATA STACKS
Confidential. Copyright © 2017 Accelerite. All rights reserved.
Technology: Massively Parallel Processing (MPP) Database
Workload: Decision Support1995 -
Data Stack 2.0
Technology: Hadoop –Distributed Storage & Flexible Compute
Workload: Advanced & Diverse Analytics2007 -
Data Stack 3.0
Technology: Relational Databases
Workload: Online Transaction Processing (OLTP)
Architecture Evolution:• Single Tier• Client-Server• Three-tier• Web Application
Data Stack 1.0
1980 -
DATA STACK 3.0: WHAT IS BIG DATA?
Confidential. Copyright © 2017 Accelerite. All rights reserved.
Data First, Schema Later
Schema/Structure
Data
Technology: Hadoop –Distributed Storage & Flexible Compute
Workload: Advanced & Diverse Analytics2007
Data Stack 3.0
Commodity H/W & N/WDesigned for Distributed
DATA STACK 3.0: WHAT IS BIG DATA GOOD AT ?
Confidential. Copyright © 2017 Accelerite. All rights reserved.
Analyze both Structured and Unstructured Rapid Analytics Building
Multi Structured Data: Data Stack 3.0 helps you analyze huge amounts of unstructured data
Variety of Formats: Data Stack 3.0 enables you to derive insights hidden within disparate data spread across systems and formats
Data Stack 3.0
2007
Handling LargeVolumes of Data
Confidential. Copyright © 2017 Accelerite. All rights reserved.
Fraud Detection
Security/Risk Analytics
Internet of Things
Pricing Analytics
Customer 360
KEY USE CASES FOR BIG DATA
Source: OREILLY, The Big Data Market by Aman Naimat
© 2016 Accelerite. All Rights Reserved. Confidential. Copyright © 2017 Accelerite. All rights reserved.
Challenges in
Big Data
BIG DATA ANALYTICS IS HARD …
Confidential. Copyright © 2017 Accelerite. All rights reserved.
These are your choices…
INFORMATION FLOW IN AN ORGANIZATION
Confidential. Copyright © 2017 Accelerite. All rights reserved.
Preferred ToolsUsers in an OrgKey Questions
Can you analyze theimpact due to new XXX ?
Can you generate report that looks like YYY ?
Can you reshape the data to With this additional metrics ZZ ?
Can you please collect data from newsystems ?
PlatformDeveloper
Business Users
Data Analyst
BI Developer
ETL Developer
CHALLENGES IN BIG DATA ANALYTICS!
Multiple Technologies
• Varying abstractions
• Duplication of modelling
• Hard to manage lifecycle
Evolving Landscape
• Open source technologies die quickly and painfully
• Problem of plenty – Too many databases/tools difficult to keep in sync
• Slightly different approach towards things
Skills Gap
• Extremely hard to find qualified Big Data engineers
• Poor collaboration, makes it hard for a group to work together
• Complexity of technology – Half knowledge is more dangerous than none
Confidential. Copyright © 2017 Accelerite. All rights reserved.
© 2016 Accelerite. All Rights Reserved. Confidential. Copyright © 2017 Accelerite. All rights reserved.
Solution
WHAT IS SHAREINSIGHTS?
Confidential. Copyright © 2017 Accelerite. All rights reserved.
A tool for doing end-to-end analytics.
A single tool for doing data visualization, data processing, advanced analytics, data ingestion and storage (end-to-end).
INFORMATION FLOW IN AN ORGANIZATION
Confidential. Copyright © 2017 Accelerite. All rights reserved.
ToolsUsers in an OrgKey Questions
Can you analyze theimpact due to new XXX ?
Can you generate report that looks like YYY ?
Can you reshape the data to With this additional metrics ZZ ?
Can you please collect data from newsystems ?
PlatformDeveloper
Business Users
Data Analyst
BI Developer
ETL Developer
A single tool for data analysts.
BusinessTeam
TechnologyTeam
Hadoop InfrastructureCloud Sources
Non-Relational Sources
Data Warehouse
SQL Code
Real Time Streams
Code
VisualizationLayer
Designer: Representing End-to-End Analytics
1
Execution for End-to-End Analytics
2
WHAT IS SHAREINSIGHTS?
Confidential. Copyright © 2017 Accelerite. All rights reserved.
BusinessTeam
TechnologyTeam
Hadoop InfrastructureCloud Sources
Non-Relational Sources
Data Warehouse
SQL Code
Real Time Streams
Code
VisualizationLayer
Designer: Representing End-to-End Analytics
Execution for End-to-End Analytics
Analytics Representation
Language
Code Generation Engine
Library of Tasks, Widgets & Connectors
1 2 3
Confidential. Copyright © 2017 Accelerite. All rights reserved.
SHAREINSIGHTS: UNDER THE HOOD
Rapid Analytics Development
Managed Lifecycle
Diversity of Analytics
Future Proofing
KEY VALUE PROPOSITIONS
SAMPLE CASE STUDIES
• Customer: A consortium of media, cable and technology providers.
• What: Analyze audience behavior with the help of predictive algorithms and proprietary datasets.
• Why: Quick Turnaround: & Lack of Skills
• Pricing: Subscription
Audience Analytics:
• Customer: Provider compliance and research services to providers.
• What: Do compliance reporting for providers and provide additional analytics based on the data collected..
• Why ShareInsights: Lack of Skills
• Pricing: Enterprise
Healthcare Provider Analytics
• Customer: Media Company
• What: Analyze a launch of media event and take corrective action.
• Why: Variety of Analytics & lack of skills
• Pricing: Enterprise
Campaign Analytics
© 2016 Accelerite. All Rights Reserved. Confidential. Copyright © 2017 Accelerite. All rights reserved.
Demo
© 2016 Accelerite. All Rights Reserved. Confidential. Copyright © 2017 Accelerite. All rights reserved.
Thanks!
HYBRID BIG DATA ARCHITECTURE
Hadoop Infrastructure
Cloud Sources
Non-Relational
Sources
Real Time Streams
Data
Warehouse
Visualization
Layer
Business
Team
Infrastructure
TeamTechnology
Team
Enterprise Sources
SQL Code
Code
CODE GENERATION APPROACH
IML
File1 2 43
Data
Processing
Data
Storage
Middle
Layer
Reporting
1. Hadoop: PIG
2. Hadoop: Spark
3. Spark Streaming
4. JavaScript
1.Hadoop
2. RDMBS
3. Sqlite
4. AWS S3
1.Web Service
2.ODBC
3. JDBC
ShareInsights
Compiler
1.JavaScript Engine
2.Jasper Reports
(Metadata, Consumption, Environment)
Drag & Drop UI Tool for
IML Construction
© 2016 Accelerite. All Rights Reserved.26
1. Language & SDK: The building blocks
Lan
guag
eLi
bra
ry
& S
DK
Connector SDK
• Support for CDC & Real-time
• Library of 25+ Connectors
Task SDK
• SDK in Python, R & Java.
• Easily wrap existing Lib
• Library of 100s of Tasks.
Widget SDK
• Mature SDK in JavaScript
• Library of over 70 widgets.
1) Entities:
2) Collections
3) Single Config.
1. Data 2. Task 3. Widget
4. Flows: {Tasks}5.Layout:
{Widgets}
Data ModelUnified
Security
Managed
LifecycleCentral
Deployment
3. Environment
API Driven
Collaborate & Share
On-premise & Cloud
Zero DeveloperFootprint
Unified Securityacross Layers
Manage the Lifecycle
top related