systems of intelligence: the biggest change in enterprise applications in 50 years
TRANSCRIPT
Systems of Intelligence:The Biggest Change in Enterprise Applications in 50 Years
George Gilbert@GGilbert41
Big Data Analyst
StoreE-Mail
SocialMedia IM
Where Customers Should be InvestingAnalogous to Pipelines in Systems of Record
Operational apps
Customer interactions
Agility: Speed of improving reportsweeks, months, quarters
Latency: Speed of reportingDays, weeks, months
ETL Development
OperationalData
Customer“Breadcrumbs”
Production ETL
StoreE-Mail
SocialMedia IM
Where Customers Should be InvestingSystems of Intelligence Pipelines are all About Speed and Agility
Retail
ConsumerMobile
Call Center
eCommerce
Operational apps
Customer interactions
Agility: Improving predictions seconds to days
Latency: Speed of predictionsms to seconds
Machine learning pipeline
OperationalData
Customer“Breadcrumbs”
Predictions,Recommendations
ImprovingPredictions(Machine Learning)
Intelligence
Critical new data and analytic skills are required• Accuracy of predictions (Revenue)
Modernizing SoR can accelerate the journey• Speed of predictions (latency)• Speed of improving predictions (agility)
Choice of new platform• TCO/operational complexity• Development complexity• Existing infrastructure – technology and
skills
What Trade-Offs Should Customers Consider When Deploying Systems of Intelligence
Incremental Revenue
Planning Your Customer Journey: Skills and Platform Progress
Smart Grid
Fraud prevention
Real-time loyaltyomni-channelmulti-touchpoint
Predictive model learns from and anticipates consumer in near real-time
Continuously updated prediction of energy supply, demand tunes end-point consumptionIntelligent
systems management
System learns “normal” behavior of apps and infrastructure and flags or fixes anomalies
Identify spending behavior out of the norm
ApplicationsApplications
Technology Maturity, Enterprise Capabilites
Technology Maturity, Enterprise CapabilitesTime
Big Data Platform Evolution – Simplifying Operations and Development:Storage Consolidation, API’s > OSS Products, Spark Hollowing out Hadoop
Machine Learning
SQL: Join, filter, aggregate
Streaming
HDFS-Compatible File System
Resource + Workload Manager
HBase-Compatible DB
Polyglot Data API: SQL, KV, JSON
YARN Resource Manager
HDFS, HBase
StreamingSpark, Flume, Flink, SamzaDataflow
Kafka-Like Messaging
SQLImpala, Drill, Hive, HAWQ…
Machine LearningMahout,
Spark MLlib
Hadoop 2.0 Big Data 3.0
In-Memory Storage (DB or File System)
Dev Tools Dev Tools Dev Tools
Dev Tool
Inte
grat
ed D
evop
s
Man
agem
ent,
Gove
rnan
ce
Operational Simplicity vs. Complexity:Single Pane of Glass vs. Familiar Legacy Sprawl
Development Simplicity vs. Complexity: Spreadsheet vs. Many Development Tools
Many data managers – optimal functionality(Cassandra, Aerospike, MongoDB, Neo4j…)
Single vendor data platform(Azure, AWS, Google Cloud Platform, Bluemix, Pivotal)
Single multi-purpose engine(Oracle, Spark)
Customers Must Balance Operational Simplicity and Development SimplicityRelative to Existing Skills and Infrastructure
Ope
ratio
nal
Sim
plic
ity
Development Simplicity
Hadoop ecosystem(Hortonworks, Cloudera, MapR)
SaaS: AirBnB, Uber, Commerce
Fortune 500
Mainstream
Pro: Optimal functionalityCon: ComplexityCustomer sweet spot• Leading-edge Internet-
centric companies• Netflix, Uber, ad-tech,
gaming, ecommerce
Many Data Managers – Optimal Functionality: Highest Development and Operational Complexity
Many optimized data managers(Cassandra, Aerospike, MongoDB, Neo4j…)
Single vendor data platform(Azure, AWS, Google Cloud Platform, Bluemix, Pivotal)
Single multi-purpose engine(Oracle, Spark)
Ope
ratio
nal
Sim
plic
ity
Development Simplicity
Hadoop ecosystem(Hortonworks, Cloudera, MapR)
Hadoop 2.0
Big Data 3.0
Customers Must Balance Operational Simplicity and Development SimplicityRelative to Existing Skills and Infrastructure
Critical new data and analytic skills are required• Accuracy of predictions
Modernizing SoR can accelerate the journey• Speed of predictions (latency)• Speed of improving predictions (agility)
Choice of new platform• TCO/operational complexity• Development complexity• Existing infrastructure – technology and
skills
Systems of Intelligence P&L Statement: Designing with “Budgetary” Constraints
Technology maturity:• lowers cost • opens new application possibilities
Incremental Revenue