SlideShare a Scribd company logo
1 of 36
How real-time data processing is used for
application in customer experience?
by Christos Hadjinikolis | SeniorML Engineer| Data Reply UK
Data ReplyUK
Data Reply is the Reply group company offering a broad range of analytics and data
processing services.
• Data Science
• Data Engineering
• Visit our website:https://www.reply.com/data-reply/en/
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 2
Agenda
• An Introduction to Streaming
• Customer Experience & Real-time Analytics
• Intro
• Approaches
• Use Cases
• Pain points
• Why Flink?
• Capabilities
• TheOpen SourceCommunity
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 3
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 4
Two maindrivespush us
towardsevolution!
1. TheMarket incentives
2. Technologyrequirements
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 5
Two maindrivespush us
towardsevolution!
1. TheMarket incentives
Incentives aim to provide
value for money and
contribute towards some
sort of success with
regards to certain goals.
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 6
Two maindrivespush us
towardsevolution!
1. TheMarket incentives
2. Technologyrequirements
The second is really a
direct consequence of the
first. In our effort to
increase our efficiency
and the effectiveness of
our solutions we
constantly improve the
tools we use.
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 7
…we come up withsolutions thatlead to
the emergence of new paradigms
which oftenend up defininghow the
market evolves!
“
“
…let’s take a step
back
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 9
Traditional Data
Infrastructures
1. TransactionalProcessing
*from Stream Processing with Apache Flink, by F. Huaske & V. Kalavri
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 10
Traditional Data
Infrastructures
1. TransactionalProcessing
• Microservices
*from Stream Processing with Apache Flink, by F. Huaske & V. Kalavri
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 11
Traditional Data
Infrastructures
1. TransactionalProcessing
• Microservices
2. Analytical Processing
*from Stream Processing with Apache Flink, by F. Huaske & V. Kalavri
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 12
Statefull Processing
• Virtually all data is createdas continuous streams
of events. (Event logs)
• Not just record-at-a-time transformations (local
state)
• An Example
*from Stream Processing with Apache Flink, by F. Huaske & V. Kalavri
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 13
Statefull Processing
• Event-Driven Applications
• Real-timerecommendations(e.g., forrecommending
productswhile customersbrowsearetailer’swebsite)
• Patterndetection or complexevent processing (e.g.,
forfrauddetectionin credit cardtransactions)
• Anomalydetection(e.g., todetect attemptstointrudea
computernetwork)
*from Stream Processing with Apache Flink, by F. Huaske & V. Kalavri
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 14
Statefull Processing
• Streaming analytics
• Monitoringthequalityof customer
communications
• Analyzinguserbehavior in mobileapplications
• Ad-hocanalysisoflive datain consumertechnology
*from Stream Processing with Apache Flink, by F. Huaske & V. Kalavri
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 15
Customer Experience
• Findand applythe best possibleway to
buildloyalty, reduce churn and generate
incrementalrevenue.
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 16
Timely Visibility is key
• Reactive/ad-hoc marketing
• Improved awareness
• Immediately and automatically collect, correlate,
analyzeand act on customerdata in diverse formats
across systems within secondsofoccurrence.
• Addressproblems before theycyclethrough and
negatively affect the customerexperience
So, how can stream processing
technologies help with that?
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 18
Real-time situational
awareness
• Problems may occurat anytime:
• customerservice satisfaction;
• networkproblems;
• fraudulent activityandmore.
• Being able to detect these issuesas they’rehappening
protects the customer—and the broaderorganization—
from the fall-out.
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 19
Real-time 1:1 marketing
• Offer micro-targeted, personalized offers that
both enhance the customer experiencethrough
providing highly relevant, timely value and drive
incremental revenuethrough upsell and cross-sell
opportunities.
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 20
Real-time fraud detection and
prevention
• Speed is key
• It’s critically important that companies are ableto notonly to detect
and stop fraud as it’s happening—down to the secondof
occurrence—but also to leverage predictive analytics to preventit
from happening in the first place..
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 21
Real-time Recommendations
Gather customer insights can be leveragedto:
• Predict customerneeds
• Predict customerchurn
• Adjust product strategy
• Adjust marketing strategy
…let’s look at some
use-cases
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 23
Streaming Media Industry
• FragmentedIndustry
• RabbitHoleproblem!
• Users get lost
• Difficult to choose from the available content!
• How can companies make a difference?
• AdvancedUser Search
• RecommenderSystems
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 24
NetFlix & YouTube
• Monetization strategy:
• Premium Memberships & Advertising
Challenge:
• Makerecommendationsfromavery large corpusofvideos,while being
certainthatthesmall numberof videosthatwill appearonone’s device
are personalised andengagingforthe user.
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 25
Solution
• Collaborative Filtering
• Technology:2 NNLayers
Video Corpus
Candidate
Generation
Ranking
x106
x102
Video Features
x10
User history
…why Flink then?
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 27
Scalability& Durability
• Flink is a distributed processing framework
• Millisecond latencies while processing millions of events
persecond.
• Exactly-once state consistency guarantees.
• Ability to run streaming applications 24/7
• Fault Tolerant
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 28
IoTExplosion
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 29
5G estimated to reach 1.5 billion
subscriptions in 2024
1.5
billion
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 30
Some very interesting libs
• CEP:Complex Event Processing
Combines datafrommultiple sourcestoinfer eventsorpatterns
thatsuggest morecomplicatedcircumstances.
• SQL on Streams
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 31
Smooth Model Serving
• Stream serialised models into your processing engine
just like any other message
• Assemble the model andstart using it without
disruptions!
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 32
Trusted bymajor companies
• Alibaba
• Ebay
• Uber
• AWS
• Huawei
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 33
Open Source Community
• Supported by a vast and active community
• Fast and agile evolution
• Transparency
• Reliability
• Cost effective
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 34
Flink Meetup in London
Community
• Around since 2016
• 600+ members
• O’Reily Partners
• Wehost events regularly (once a month)
Data ReplyUK
Data Reply is the Reply group company offering a broad range of analytics and data
processing services.
• Data Science
• Data Engineering
• Visit our website:https://www.reply.com/data-reply/en/
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 35
18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 36
Thanks

More Related Content

What's hot

Big data, big deal, Acxiom
Big data, big deal, Acxiom  Big data, big deal, Acxiom
Big data, big deal, Acxiom Internet World
 
18th Athens Big Data Meetup - 1st Talk - Timeseries Forecasting as a Service
18th Athens Big Data Meetup - 1st Talk - Timeseries Forecasting as a Service18th Athens Big Data Meetup - 1st Talk - Timeseries Forecasting as a Service
18th Athens Big Data Meetup - 1st Talk - Timeseries Forecasting as a ServiceAthens Big Data
 
Top 5 Deep Learning and AI Stories - September 14, 2018
Top 5 Deep Learning and AI Stories - September 14, 2018Top 5 Deep Learning and AI Stories - September 14, 2018
Top 5 Deep Learning and AI Stories - September 14, 2018NVIDIA
 
IDC Portugal | Como Libertar os Seus Dados com Virtualização de Dados
IDC Portugal | Como Libertar os Seus Dados com Virtualização de DadosIDC Portugal | Como Libertar os Seus Dados com Virtualização de Dados
IDC Portugal | Como Libertar os Seus Dados com Virtualização de DadosDenodo
 
KnowNow Introduction to application
KnowNow Introduction to applicationKnowNow Introduction to application
KnowNow Introduction to applicationDavid Patterson
 
European Utility Week 2015: Speed vs Quality in Smart Grid Analytics
European Utility Week 2015: Speed vs Quality in Smart Grid AnalyticsEuropean Utility Week 2015: Speed vs Quality in Smart Grid Analytics
European Utility Week 2015: Speed vs Quality in Smart Grid AnalyticsOMNETRIC
 
The AI Opportunity in Federal - Key Highlights from GTC DC 2018
The AI Opportunity in Federal - Key Highlights from GTC DC 2018The AI Opportunity in Federal - Key Highlights from GTC DC 2018
The AI Opportunity in Federal - Key Highlights from GTC DC 2018NVIDIA
 
Research Problem Presentation - Research in Supply Chain Digital Twins
Research Problem Presentation - Research in Supply Chain Digital TwinsResearch Problem Presentation - Research in Supply Chain Digital Twins
Research Problem Presentation - Research in Supply Chain Digital TwinsArwa Abougharib
 
Celebrating and Supporting the Medical Imaging Community
Celebrating and Supporting the Medical Imaging CommunityCelebrating and Supporting the Medical Imaging Community
Celebrating and Supporting the Medical Imaging CommunityNVIDIA
 
Pyramid Pressefrühstück (Präsentation Intel) München
Pyramid Pressefrühstück (Präsentation Intel) MünchenPyramid Pressefrühstück (Präsentation Intel) München
Pyramid Pressefrühstück (Präsentation Intel) MünchenPyramid Computer GmbH
 
2016 IDC Pan-European Utilities Summit: Open for Business
2016 IDC Pan-European Utilities Summit: Open for Business2016 IDC Pan-European Utilities Summit: Open for Business
2016 IDC Pan-European Utilities Summit: Open for BusinessOMNETRIC
 
Dltop5 2017-10-06v2-171006205635 (1)
Dltop5 2017-10-06v2-171006205635 (1)Dltop5 2017-10-06v2-171006205635 (1)
Dltop5 2017-10-06v2-171006205635 (1)Cemil Yigit
 
Advancing Medical Imaging with Deep Learning
Advancing Medical Imaging with Deep LearningAdvancing Medical Imaging with Deep Learning
Advancing Medical Imaging with Deep LearningNVIDIA
 

What's hot (13)

Big data, big deal, Acxiom
Big data, big deal, Acxiom  Big data, big deal, Acxiom
Big data, big deal, Acxiom
 
18th Athens Big Data Meetup - 1st Talk - Timeseries Forecasting as a Service
18th Athens Big Data Meetup - 1st Talk - Timeseries Forecasting as a Service18th Athens Big Data Meetup - 1st Talk - Timeseries Forecasting as a Service
18th Athens Big Data Meetup - 1st Talk - Timeseries Forecasting as a Service
 
Top 5 Deep Learning and AI Stories - September 14, 2018
Top 5 Deep Learning and AI Stories - September 14, 2018Top 5 Deep Learning and AI Stories - September 14, 2018
Top 5 Deep Learning and AI Stories - September 14, 2018
 
IDC Portugal | Como Libertar os Seus Dados com Virtualização de Dados
IDC Portugal | Como Libertar os Seus Dados com Virtualização de DadosIDC Portugal | Como Libertar os Seus Dados com Virtualização de Dados
IDC Portugal | Como Libertar os Seus Dados com Virtualização de Dados
 
KnowNow Introduction to application
KnowNow Introduction to applicationKnowNow Introduction to application
KnowNow Introduction to application
 
European Utility Week 2015: Speed vs Quality in Smart Grid Analytics
European Utility Week 2015: Speed vs Quality in Smart Grid AnalyticsEuropean Utility Week 2015: Speed vs Quality in Smart Grid Analytics
European Utility Week 2015: Speed vs Quality in Smart Grid Analytics
 
The AI Opportunity in Federal - Key Highlights from GTC DC 2018
The AI Opportunity in Federal - Key Highlights from GTC DC 2018The AI Opportunity in Federal - Key Highlights from GTC DC 2018
The AI Opportunity in Federal - Key Highlights from GTC DC 2018
 
Research Problem Presentation - Research in Supply Chain Digital Twins
Research Problem Presentation - Research in Supply Chain Digital TwinsResearch Problem Presentation - Research in Supply Chain Digital Twins
Research Problem Presentation - Research in Supply Chain Digital Twins
 
Celebrating and Supporting the Medical Imaging Community
Celebrating and Supporting the Medical Imaging CommunityCelebrating and Supporting the Medical Imaging Community
Celebrating and Supporting the Medical Imaging Community
 
Pyramid Pressefrühstück (Präsentation Intel) München
Pyramid Pressefrühstück (Präsentation Intel) MünchenPyramid Pressefrühstück (Präsentation Intel) München
Pyramid Pressefrühstück (Präsentation Intel) München
 
2016 IDC Pan-European Utilities Summit: Open for Business
2016 IDC Pan-European Utilities Summit: Open for Business2016 IDC Pan-European Utilities Summit: Open for Business
2016 IDC Pan-European Utilities Summit: Open for Business
 
Dltop5 2017-10-06v2-171006205635 (1)
Dltop5 2017-10-06v2-171006205635 (1)Dltop5 2017-10-06v2-171006205635 (1)
Dltop5 2017-10-06v2-171006205635 (1)
 
Advancing Medical Imaging with Deep Learning
Advancing Medical Imaging with Deep LearningAdvancing Medical Imaging with Deep Learning
Advancing Medical Imaging with Deep Learning
 

Similar to Big Data London Meetup on Customer Experience

Predictive vs Prescriptive Analytics
Predictive vs Prescriptive AnalyticsPredictive vs Prescriptive Analytics
Predictive vs Prescriptive AnalyticsDATAVERSITY
 
Herding Cats in the Digital World
Herding Cats in the Digital WorldHerding Cats in the Digital World
Herding Cats in the Digital WorldCapgemini
 
The value of a connected factory
The value of a connected factoryThe value of a connected factory
The value of a connected factoryCroonwolter&dros
 
Scaling Your Enterprise With Data Science
Scaling Your Enterprise With Data ScienceScaling Your Enterprise With Data Science
Scaling Your Enterprise With Data ScienceSuperFluid Labs
 
2019 Global Report Digital transformation | Schneider Electric
2019 Global  Report Digital transformation | Schneider Electric2019 Global  Report Digital transformation | Schneider Electric
2019 Global Report Digital transformation | Schneider ElectricMassimo Talia
 
Siecap Advisory Automation & Supply Chain Trends
Siecap Advisory Automation & Supply Chain TrendsSiecap Advisory Automation & Supply Chain Trends
Siecap Advisory Automation & Supply Chain TrendsGeoffrey Knowles
 
Data Center of the Future: Designing a modernized, high performance computing...
Data Center of the Future: Designing a modernized, high performance computing...Data Center of the Future: Designing a modernized, high performance computing...
Data Center of the Future: Designing a modernized, high performance computing...Capgemini
 
20131202 Value of supply chain analytics for the electronics industry - a st...
20131202  Value of supply chain analytics for the electronics industry - a st...20131202  Value of supply chain analytics for the electronics industry - a st...
20131202 Value of supply chain analytics for the electronics industry - a st...Thorsten Schroeer
 
CWIN17 san francisco-shawn kelly-iot business value
CWIN17 san francisco-shawn kelly-iot business valueCWIN17 san francisco-shawn kelly-iot business value
CWIN17 san francisco-shawn kelly-iot business valueCapgemini
 
Swoc21 Feb08 Amig
Swoc21 Feb08 AmigSwoc21 Feb08 Amig
Swoc21 Feb08 Amiglatha_only
 
UtiliVATION - Utility Innovation - Indigo Advisory Group
UtiliVATION - Utility Innovation - Indigo Advisory GroupUtiliVATION - Utility Innovation - Indigo Advisory Group
UtiliVATION - Utility Innovation - Indigo Advisory GroupIndigo Advisory Group
 
Analytics ROI Best Practices
Analytics ROI Best PracticesAnalytics ROI Best Practices
Analytics ROI Best PracticesDATAVERSITY
 
Augmented OLAP Analytics for Big Data
Augmented OLAP Analytics for Big DataAugmented OLAP Analytics for Big Data
Augmented OLAP Analytics for Big DataTyler Wishnoff
 
Augmented OLAP for Big Data
Augmented OLAP for Big DataAugmented OLAP for Big Data
Augmented OLAP for Big DataLuke Han
 
Capgemini ses - smart analytics accelerates the realization of value for ut...
Capgemini   ses - smart analytics accelerates the realization of value for ut...Capgemini   ses - smart analytics accelerates the realization of value for ut...
Capgemini ses - smart analytics accelerates the realization of value for ut...Gord Reynolds
 
Cloud Billing: Enabling consumers for pay for what they use
Cloud Billing: Enabling consumers for pay for what they useCloud Billing: Enabling consumers for pay for what they use
Cloud Billing: Enabling consumers for pay for what they useEduardo Mendez Polo
 

Similar to Big Data London Meetup on Customer Experience (20)

Digital at WMG, Simon Buckingham
Digital at WMG, Simon BuckinghamDigital at WMG, Simon Buckingham
Digital at WMG, Simon Buckingham
 
Predictive vs Prescriptive Analytics
Predictive vs Prescriptive AnalyticsPredictive vs Prescriptive Analytics
Predictive vs Prescriptive Analytics
 
Herding Cats in the Digital World
Herding Cats in the Digital WorldHerding Cats in the Digital World
Herding Cats in the Digital World
 
The value of a connected factory
The value of a connected factoryThe value of a connected factory
The value of a connected factory
 
Scaling Your Enterprise With Data Science
Scaling Your Enterprise With Data ScienceScaling Your Enterprise With Data Science
Scaling Your Enterprise With Data Science
 
2019 Global Report Digital transformation | Schneider Electric
2019 Global  Report Digital transformation | Schneider Electric2019 Global  Report Digital transformation | Schneider Electric
2019 Global Report Digital transformation | Schneider Electric
 
Ark Product and Process Design V1
Ark Product and Process Design V1Ark Product and Process Design V1
Ark Product and Process Design V1
 
DASA Security Showcase - DASA Presentation
DASA Security Showcase - DASA PresentationDASA Security Showcase - DASA Presentation
DASA Security Showcase - DASA Presentation
 
Siecap Advisory Automation & Supply Chain Trends
Siecap Advisory Automation & Supply Chain TrendsSiecap Advisory Automation & Supply Chain Trends
Siecap Advisory Automation & Supply Chain Trends
 
Data Center of the Future: Designing a modernized, high performance computing...
Data Center of the Future: Designing a modernized, high performance computing...Data Center of the Future: Designing a modernized, high performance computing...
Data Center of the Future: Designing a modernized, high performance computing...
 
20131202 Value of supply chain analytics for the electronics industry - a st...
20131202  Value of supply chain analytics for the electronics industry - a st...20131202  Value of supply chain analytics for the electronics industry - a st...
20131202 Value of supply chain analytics for the electronics industry - a st...
 
CWIN17 san francisco-shawn kelly-iot business value
CWIN17 san francisco-shawn kelly-iot business valueCWIN17 san francisco-shawn kelly-iot business value
CWIN17 san francisco-shawn kelly-iot business value
 
Can you afford not to consider sustainable ICT?
Can you afford not to consider sustainable ICT?Can you afford not to consider sustainable ICT?
Can you afford not to consider sustainable ICT?
 
Swoc21 Feb08 Amig
Swoc21 Feb08 AmigSwoc21 Feb08 Amig
Swoc21 Feb08 Amig
 
UtiliVATION - Utility Innovation - Indigo Advisory Group
UtiliVATION - Utility Innovation - Indigo Advisory GroupUtiliVATION - Utility Innovation - Indigo Advisory Group
UtiliVATION - Utility Innovation - Indigo Advisory Group
 
Analytics ROI Best Practices
Analytics ROI Best PracticesAnalytics ROI Best Practices
Analytics ROI Best Practices
 
Augmented OLAP Analytics for Big Data
Augmented OLAP Analytics for Big DataAugmented OLAP Analytics for Big Data
Augmented OLAP Analytics for Big Data
 
Augmented OLAP for Big Data
Augmented OLAP for Big DataAugmented OLAP for Big Data
Augmented OLAP for Big Data
 
Capgemini ses - smart analytics accelerates the realization of value for ut...
Capgemini   ses - smart analytics accelerates the realization of value for ut...Capgemini   ses - smart analytics accelerates the realization of value for ut...
Capgemini ses - smart analytics accelerates the realization of value for ut...
 
Cloud Billing: Enabling consumers for pay for what they use
Cloud Billing: Enabling consumers for pay for what they useCloud Billing: Enabling consumers for pay for what they use
Cloud Billing: Enabling consumers for pay for what they use
 

Recently uploaded

OSVC_Meta-Data based Simulation Automation to overcome Verification Challenge...
OSVC_Meta-Data based Simulation Automation to overcome Verification Challenge...OSVC_Meta-Data based Simulation Automation to overcome Verification Challenge...
OSVC_Meta-Data based Simulation Automation to overcome Verification Challenge...Soham Mondal
 
Call Girls Delhi {Jodhpur} 9711199012 high profile service
Call Girls Delhi {Jodhpur} 9711199012 high profile serviceCall Girls Delhi {Jodhpur} 9711199012 high profile service
Call Girls Delhi {Jodhpur} 9711199012 high profile servicerehmti665
 
HARMONY IN THE NATURE AND EXISTENCE - Unit-IV
HARMONY IN THE NATURE AND EXISTENCE - Unit-IVHARMONY IN THE NATURE AND EXISTENCE - Unit-IV
HARMONY IN THE NATURE AND EXISTENCE - Unit-IVRajaP95
 
MANUFACTURING PROCESS-II UNIT-2 LATHE MACHINE
MANUFACTURING PROCESS-II UNIT-2 LATHE MACHINEMANUFACTURING PROCESS-II UNIT-2 LATHE MACHINE
MANUFACTURING PROCESS-II UNIT-2 LATHE MACHINESIVASHANKAR N
 
Software Development Life Cycle By Team Orange (Dept. of Pharmacy)
Software Development Life Cycle By  Team Orange (Dept. of Pharmacy)Software Development Life Cycle By  Team Orange (Dept. of Pharmacy)
Software Development Life Cycle By Team Orange (Dept. of Pharmacy)Suman Mia
 
Decoding Kotlin - Your guide to solving the mysterious in Kotlin.pptx
Decoding Kotlin - Your guide to solving the mysterious in Kotlin.pptxDecoding Kotlin - Your guide to solving the mysterious in Kotlin.pptx
Decoding Kotlin - Your guide to solving the mysterious in Kotlin.pptxJoão Esperancinha
 
Introduction to IEEE STANDARDS and its different types.pptx
Introduction to IEEE STANDARDS and its different types.pptxIntroduction to IEEE STANDARDS and its different types.pptx
Introduction to IEEE STANDARDS and its different types.pptxupamatechverse
 
Call Girls Service Nagpur Tanvi Call 7001035870 Meet With Nagpur Escorts
Call Girls Service Nagpur Tanvi Call 7001035870 Meet With Nagpur EscortsCall Girls Service Nagpur Tanvi Call 7001035870 Meet With Nagpur Escorts
Call Girls Service Nagpur Tanvi Call 7001035870 Meet With Nagpur EscortsCall Girls in Nagpur High Profile
 
Extrusion Processes and Their Limitations
Extrusion Processes and Their LimitationsExtrusion Processes and Their Limitations
Extrusion Processes and Their Limitations120cr0395
 
the ladakh protest in leh ladakh 2024 sonam wangchuk.pptx
the ladakh protest in leh ladakh 2024 sonam wangchuk.pptxthe ladakh protest in leh ladakh 2024 sonam wangchuk.pptx
the ladakh protest in leh ladakh 2024 sonam wangchuk.pptxhumanexperienceaaa
 
(RIA) Call Girls Bhosari ( 7001035870 ) HI-Fi Pune Escorts Service
(RIA) Call Girls Bhosari ( 7001035870 ) HI-Fi Pune Escorts Service(RIA) Call Girls Bhosari ( 7001035870 ) HI-Fi Pune Escorts Service
(RIA) Call Girls Bhosari ( 7001035870 ) HI-Fi Pune Escorts Serviceranjana rawat
 
(ANJALI) Dange Chowk Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...
(ANJALI) Dange Chowk Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...(ANJALI) Dange Chowk Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...
(ANJALI) Dange Chowk Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...ranjana rawat
 
MANUFACTURING PROCESS-II UNIT-5 NC MACHINE TOOLS
MANUFACTURING PROCESS-II UNIT-5 NC MACHINE TOOLSMANUFACTURING PROCESS-II UNIT-5 NC MACHINE TOOLS
MANUFACTURING PROCESS-II UNIT-5 NC MACHINE TOOLSSIVASHANKAR N
 
Introduction to Multiple Access Protocol.pptx
Introduction to Multiple Access Protocol.pptxIntroduction to Multiple Access Protocol.pptx
Introduction to Multiple Access Protocol.pptxupamatechverse
 
(MEERA) Dapodi Call Girls Just Call 7001035870 [ Cash on Delivery ] Pune Escorts
(MEERA) Dapodi Call Girls Just Call 7001035870 [ Cash on Delivery ] Pune Escorts(MEERA) Dapodi Call Girls Just Call 7001035870 [ Cash on Delivery ] Pune Escorts
(MEERA) Dapodi Call Girls Just Call 7001035870 [ Cash on Delivery ] Pune Escortsranjana rawat
 
High Profile Call Girls Nagpur Meera Call 7001035870 Meet With Nagpur Escorts
High Profile Call Girls Nagpur Meera Call 7001035870 Meet With Nagpur EscortsHigh Profile Call Girls Nagpur Meera Call 7001035870 Meet With Nagpur Escorts
High Profile Call Girls Nagpur Meera Call 7001035870 Meet With Nagpur EscortsCall Girls in Nagpur High Profile
 
VIP Call Girls Service Kondapur Hyderabad Call +91-8250192130
VIP Call Girls Service Kondapur Hyderabad Call +91-8250192130VIP Call Girls Service Kondapur Hyderabad Call +91-8250192130
VIP Call Girls Service Kondapur Hyderabad Call +91-8250192130Suhani Kapoor
 
(ANVI) Koregaon Park Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...
(ANVI) Koregaon Park Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...(ANVI) Koregaon Park Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...
(ANVI) Koregaon Park Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...ranjana rawat
 

Recently uploaded (20)

OSVC_Meta-Data based Simulation Automation to overcome Verification Challenge...
OSVC_Meta-Data based Simulation Automation to overcome Verification Challenge...OSVC_Meta-Data based Simulation Automation to overcome Verification Challenge...
OSVC_Meta-Data based Simulation Automation to overcome Verification Challenge...
 
Call Girls Delhi {Jodhpur} 9711199012 high profile service
Call Girls Delhi {Jodhpur} 9711199012 high profile serviceCall Girls Delhi {Jodhpur} 9711199012 high profile service
Call Girls Delhi {Jodhpur} 9711199012 high profile service
 
HARMONY IN THE NATURE AND EXISTENCE - Unit-IV
HARMONY IN THE NATURE AND EXISTENCE - Unit-IVHARMONY IN THE NATURE AND EXISTENCE - Unit-IV
HARMONY IN THE NATURE AND EXISTENCE - Unit-IV
 
Exploring_Network_Security_with_JA3_by_Rakesh Seal.pptx
Exploring_Network_Security_with_JA3_by_Rakesh Seal.pptxExploring_Network_Security_with_JA3_by_Rakesh Seal.pptx
Exploring_Network_Security_with_JA3_by_Rakesh Seal.pptx
 
MANUFACTURING PROCESS-II UNIT-2 LATHE MACHINE
MANUFACTURING PROCESS-II UNIT-2 LATHE MACHINEMANUFACTURING PROCESS-II UNIT-2 LATHE MACHINE
MANUFACTURING PROCESS-II UNIT-2 LATHE MACHINE
 
Software Development Life Cycle By Team Orange (Dept. of Pharmacy)
Software Development Life Cycle By  Team Orange (Dept. of Pharmacy)Software Development Life Cycle By  Team Orange (Dept. of Pharmacy)
Software Development Life Cycle By Team Orange (Dept. of Pharmacy)
 
Decoding Kotlin - Your guide to solving the mysterious in Kotlin.pptx
Decoding Kotlin - Your guide to solving the mysterious in Kotlin.pptxDecoding Kotlin - Your guide to solving the mysterious in Kotlin.pptx
Decoding Kotlin - Your guide to solving the mysterious in Kotlin.pptx
 
Introduction to IEEE STANDARDS and its different types.pptx
Introduction to IEEE STANDARDS and its different types.pptxIntroduction to IEEE STANDARDS and its different types.pptx
Introduction to IEEE STANDARDS and its different types.pptx
 
Call Girls Service Nagpur Tanvi Call 7001035870 Meet With Nagpur Escorts
Call Girls Service Nagpur Tanvi Call 7001035870 Meet With Nagpur EscortsCall Girls Service Nagpur Tanvi Call 7001035870 Meet With Nagpur Escorts
Call Girls Service Nagpur Tanvi Call 7001035870 Meet With Nagpur Escorts
 
Extrusion Processes and Their Limitations
Extrusion Processes and Their LimitationsExtrusion Processes and Their Limitations
Extrusion Processes and Their Limitations
 
the ladakh protest in leh ladakh 2024 sonam wangchuk.pptx
the ladakh protest in leh ladakh 2024 sonam wangchuk.pptxthe ladakh protest in leh ladakh 2024 sonam wangchuk.pptx
the ladakh protest in leh ladakh 2024 sonam wangchuk.pptx
 
(RIA) Call Girls Bhosari ( 7001035870 ) HI-Fi Pune Escorts Service
(RIA) Call Girls Bhosari ( 7001035870 ) HI-Fi Pune Escorts Service(RIA) Call Girls Bhosari ( 7001035870 ) HI-Fi Pune Escorts Service
(RIA) Call Girls Bhosari ( 7001035870 ) HI-Fi Pune Escorts Service
 
(ANJALI) Dange Chowk Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...
(ANJALI) Dange Chowk Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...(ANJALI) Dange Chowk Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...
(ANJALI) Dange Chowk Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...
 
MANUFACTURING PROCESS-II UNIT-5 NC MACHINE TOOLS
MANUFACTURING PROCESS-II UNIT-5 NC MACHINE TOOLSMANUFACTURING PROCESS-II UNIT-5 NC MACHINE TOOLS
MANUFACTURING PROCESS-II UNIT-5 NC MACHINE TOOLS
 
Introduction to Multiple Access Protocol.pptx
Introduction to Multiple Access Protocol.pptxIntroduction to Multiple Access Protocol.pptx
Introduction to Multiple Access Protocol.pptx
 
(MEERA) Dapodi Call Girls Just Call 7001035870 [ Cash on Delivery ] Pune Escorts
(MEERA) Dapodi Call Girls Just Call 7001035870 [ Cash on Delivery ] Pune Escorts(MEERA) Dapodi Call Girls Just Call 7001035870 [ Cash on Delivery ] Pune Escorts
(MEERA) Dapodi Call Girls Just Call 7001035870 [ Cash on Delivery ] Pune Escorts
 
High Profile Call Girls Nagpur Meera Call 7001035870 Meet With Nagpur Escorts
High Profile Call Girls Nagpur Meera Call 7001035870 Meet With Nagpur EscortsHigh Profile Call Girls Nagpur Meera Call 7001035870 Meet With Nagpur Escorts
High Profile Call Girls Nagpur Meera Call 7001035870 Meet With Nagpur Escorts
 
VIP Call Girls Service Kondapur Hyderabad Call +91-8250192130
VIP Call Girls Service Kondapur Hyderabad Call +91-8250192130VIP Call Girls Service Kondapur Hyderabad Call +91-8250192130
VIP Call Girls Service Kondapur Hyderabad Call +91-8250192130
 
(ANVI) Koregaon Park Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...
(ANVI) Koregaon Park Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...(ANVI) Koregaon Park Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...
(ANVI) Koregaon Park Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...
 
Roadmap to Membership of RICS - Pathways and Routes
Roadmap to Membership of RICS - Pathways and RoutesRoadmap to Membership of RICS - Pathways and Routes
Roadmap to Membership of RICS - Pathways and Routes
 

Big Data London Meetup on Customer Experience

  • 1. How real-time data processing is used for application in customer experience? by Christos Hadjinikolis | SeniorML Engineer| Data Reply UK
  • 2. Data ReplyUK Data Reply is the Reply group company offering a broad range of analytics and data processing services. • Data Science • Data Engineering • Visit our website:https://www.reply.com/data-reply/en/ 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 2
  • 3. Agenda • An Introduction to Streaming • Customer Experience & Real-time Analytics • Intro • Approaches • Use Cases • Pain points • Why Flink? • Capabilities • TheOpen SourceCommunity 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 3
  • 4. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 4 Two maindrivespush us towardsevolution! 1. TheMarket incentives 2. Technologyrequirements
  • 5. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 5 Two maindrivespush us towardsevolution! 1. TheMarket incentives Incentives aim to provide value for money and contribute towards some sort of success with regards to certain goals.
  • 6. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 6 Two maindrivespush us towardsevolution! 1. TheMarket incentives 2. Technologyrequirements The second is really a direct consequence of the first. In our effort to increase our efficiency and the effectiveness of our solutions we constantly improve the tools we use.
  • 7. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 7 …we come up withsolutions thatlead to the emergence of new paradigms which oftenend up defininghow the market evolves! “ “
  • 8. …let’s take a step back
  • 9. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 9 Traditional Data Infrastructures 1. TransactionalProcessing *from Stream Processing with Apache Flink, by F. Huaske & V. Kalavri
  • 10. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 10 Traditional Data Infrastructures 1. TransactionalProcessing • Microservices *from Stream Processing with Apache Flink, by F. Huaske & V. Kalavri
  • 11. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 11 Traditional Data Infrastructures 1. TransactionalProcessing • Microservices 2. Analytical Processing *from Stream Processing with Apache Flink, by F. Huaske & V. Kalavri
  • 12. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 12 Statefull Processing • Virtually all data is createdas continuous streams of events. (Event logs) • Not just record-at-a-time transformations (local state) • An Example *from Stream Processing with Apache Flink, by F. Huaske & V. Kalavri
  • 13. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 13 Statefull Processing • Event-Driven Applications • Real-timerecommendations(e.g., forrecommending productswhile customersbrowsearetailer’swebsite) • Patterndetection or complexevent processing (e.g., forfrauddetectionin credit cardtransactions) • Anomalydetection(e.g., todetect attemptstointrudea computernetwork) *from Stream Processing with Apache Flink, by F. Huaske & V. Kalavri
  • 14. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 14 Statefull Processing • Streaming analytics • Monitoringthequalityof customer communications • Analyzinguserbehavior in mobileapplications • Ad-hocanalysisoflive datain consumertechnology *from Stream Processing with Apache Flink, by F. Huaske & V. Kalavri
  • 15. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 15 Customer Experience • Findand applythe best possibleway to buildloyalty, reduce churn and generate incrementalrevenue.
  • 16. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 16 Timely Visibility is key • Reactive/ad-hoc marketing • Improved awareness • Immediately and automatically collect, correlate, analyzeand act on customerdata in diverse formats across systems within secondsofoccurrence. • Addressproblems before theycyclethrough and negatively affect the customerexperience
  • 17. So, how can stream processing technologies help with that?
  • 18. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 18 Real-time situational awareness • Problems may occurat anytime: • customerservice satisfaction; • networkproblems; • fraudulent activityandmore. • Being able to detect these issuesas they’rehappening protects the customer—and the broaderorganization— from the fall-out.
  • 19. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 19 Real-time 1:1 marketing • Offer micro-targeted, personalized offers that both enhance the customer experiencethrough providing highly relevant, timely value and drive incremental revenuethrough upsell and cross-sell opportunities.
  • 20. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 20 Real-time fraud detection and prevention • Speed is key • It’s critically important that companies are ableto notonly to detect and stop fraud as it’s happening—down to the secondof occurrence—but also to leverage predictive analytics to preventit from happening in the first place..
  • 21. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 21 Real-time Recommendations Gather customer insights can be leveragedto: • Predict customerneeds • Predict customerchurn • Adjust product strategy • Adjust marketing strategy
  • 22. …let’s look at some use-cases
  • 23. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 23 Streaming Media Industry • FragmentedIndustry • RabbitHoleproblem! • Users get lost • Difficult to choose from the available content! • How can companies make a difference? • AdvancedUser Search • RecommenderSystems
  • 24. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 24 NetFlix & YouTube • Monetization strategy: • Premium Memberships & Advertising Challenge: • Makerecommendationsfromavery large corpusofvideos,while being certainthatthesmall numberof videosthatwill appearonone’s device are personalised andengagingforthe user.
  • 25. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 25 Solution • Collaborative Filtering • Technology:2 NNLayers Video Corpus Candidate Generation Ranking x106 x102 Video Features x10 User history
  • 27. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 27 Scalability& Durability • Flink is a distributed processing framework • Millisecond latencies while processing millions of events persecond. • Exactly-once state consistency guarantees. • Ability to run streaming applications 24/7 • Fault Tolerant
  • 28. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 28 IoTExplosion
  • 29. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 29 5G estimated to reach 1.5 billion subscriptions in 2024 1.5 billion
  • 30. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 30 Some very interesting libs • CEP:Complex Event Processing Combines datafrommultiple sourcestoinfer eventsorpatterns thatsuggest morecomplicatedcircumstances. • SQL on Streams
  • 31. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 31 Smooth Model Serving • Stream serialised models into your processing engine just like any other message • Assemble the model andstart using it without disruptions!
  • 32. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 32 Trusted bymajor companies • Alibaba • Ebay • Uber • AWS • Huawei
  • 33. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 33 Open Source Community • Supported by a vast and active community • Fast and agile evolution • Transparency • Reliability • Cost effective
  • 34. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 34 Flink Meetup in London Community • Around since 2016 • 600+ members • O’Reily Partners • Wehost events regularly (once a month)
  • 35. Data ReplyUK Data Reply is the Reply group company offering a broad range of analytics and data processing services. • Data Science • Data Engineering • Visit our website:https://www.reply.com/data-reply/en/ 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 35
  • 36. 18/07/2019 Dr C. Hadjinikolis|SeniorML Engineer| Data ReplyUK 36 Thanks

Editor's Notes

  1. Incentives aim to provide value for money and contribute towards some sort of success with regards to certain goals. The second is really a direct consequent of the first. In our effort to increase our efficiency and the effectiveness of our solutions we constantly improve the tools we use. However, sometimes as we address technological problems and challenges, we come up with solutions that lead to the emergence of new paradigms which often end up defining how the market evolves.
  2. Incentives aim to provide value for money and contribute towards some sort of success with regards to certain goals. The second is really a direct consequent of the first. In our effort to increase our efficiency and the effectiveness of our solutions we constantly improve the tools we use. However, sometimes as we address technological problems and challenges, we come up with solutions that lead to the emergence of new paradigms which often end up defining how the market evolves.
  3. Incentives aim to provide value for money and contribute towards some sort of success with regards to certain goals. The second is really a direct consequent of the first. In our effort to increase our efficiency and the effectiveness of our solutions we constantly improve the tools we use. However, sometimes as we address technological problems and challenges, we come up with solutions that lead to the emergence of new paradigms which often end up defining how the market evolves.
  4. However, sometimes as we address technological problems and challenges, we come up with solutions that lead to the emergence of new paradigms which often end up defining how the market evolves. Streaming is one such example and in particular statefull stream processing
  5. Applications are usually connected to external services or face human users and continuously process incoming events such as orders, email, or clicks on a website. When an event is processed, an application reads its state or updates it by running transactions against the remote database system. Often, a database system serves multiple applications that sometimes access the same databases or tables.
  6. This application design can cause problems when applications need to evolve or scale. Since multiple applications might work on the same data representation or share the same infrastructure, changing the schema of a table or scaling a database system requires careful planning and a lot of effort. A recent approach to overcoming the tight bundling of applications is the microservices design pattern. Microservices are designed as small, self-contained, and independent applications. They follow the UNIX philosophy of doing a single thing and doing it well.
  7. Instead of running analytical queries directly on the transactional databases, the data is typically replicated to a data warehouse, a dedicated datastore for analytical query workloads. In order to populate a data warehouse, the data managed by the transactional database systems needs to be copied to it. The process of copying data to the data warehouse is called extract–transform–load (ETL). An ETL process extracts data from a transactional database, transforms it into a common representation that might include validation, value normalization, encoding, deduplication, and schema transformation, and finally loads it into the analytical database. 
  8. Local state access provides very good performance compared to reading and writing queries against remote datastores.
  9. Event-driven applications are stateful streaming applications that ingest event streams and process the events with application-specific business logic. Depending on the business logic, an event-driven application can trigger actions such as sending an alert or an email or write events to an outgoing event stream to be consumed by another event-driven application. Event-driven applications are an evolution of microservices. They communicate via event logs instead of REST calls and hold application data as local state instead of writing it to and reading it from an external datastore, such as a relational database or key-value store.
  10. How do batch analytics work? Depending on the scheduling intervals it may take hours or days until a data point is included in a report. To some extent, the latency can be reduced by importing data into the datastore with a data pipeline application. However, even with continuous ETL there will always be a delay until an event is processed by a query.
  11. When problems around quality service provision surface, they cycle through the customer experience at a large scale quickly and decisively.  Dropped calls, network problems, fraudulent activity and more – the ability to detect these issues as they’re happening protects the customer – and the broader organization –  from the fall-out. The problem for CSPs that aren’t leveraging streaming analytics is that they can’t detect the disruptions fast enough – within seconds – to deflect the impact on the customer base.
  12. If CSPs are able to track the activity and movements of their subscribers at an individual level in real-time, they are in a great position to offer micro-targeted, personalized offers that both enhance the customer experience through providing highly relevant, timely value and drive incremental revenue through upsell and cross-sell opportunities. For instance, if a CSP can track the movements of an individual subscriber through from cell to cell, and, based on the trajectory, predict that the subscriber is about to leave the country and turn off data service to avoid international roaming charges, the CSP can issue a timely offer for a short-term international roaming plan.  This not only provides the subscriber with a valuable service during their travels when they otherwise would have probably not used the service at all but also drives incremental revenue through upsell.
  13. A fragmented market A marketplace where there is no one company that can exert enough influence to move the industry in a particular direction.   AI Applications in the Streaming Media industry: It’s a fragmented market Rabbit hole problem: too much content, users get lost To differentiate, one needs to leverage AI in two ways: Recommender Systems and advanced User search  
  14. Monetization strategy: Premium memberships and advertising The YouTube Recommender System has to make recommendations from a very large corpus of videos (with order of magnitude being much higher that Netflix’s) while being certain that the small number of videos that will appear on one’s device are personalised and engaging for the user.
  15. Collaborative Filtering Deep Learning: 2 steps: Candidate Generation, then Ranking (A/B testing to improve)
  16. Add these numbers with the emergence of 5G
  17. The goal of complex event processing is to identify meaningful events (such as opportunities or threats) and respond to them as quickly as possible.
  18. The goal of complex event processing is to identify meaningful events (such as opportunities or threats) and respond to them as quickly as possible.
  19. The goal of complex event processing is to identify meaningful events (such as opportunities or threats) and respond to them as quickly as possible.
  20. Open source is becoming the norm!
  21. Open source is becoming the norm!
  22. Open source is becoming the norm!