Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
Big data @ magyar telekom
Budapest Data 2016
Gergely vItárIus
CLASSIC BI DEVELOPMENT & OPERATIONS
NEXT GENERATION BI TOOLS BIG DATA TECHNOLOGIES
ADVANCED / REALTIME ANALYTICS
BI innovation center HEAD OF BI INNOVATION CENTER
IT DIRECTORATE MAGYAR TELEKOM PLC.
Big data fever (2013/14)
Event based churn (teradata / marketing)
Big data fever (2013/14)
Event based churn (teradata / marketing)
Social media command center (hadoop / T-SYSTEMS)
Big data fever (2013/14)
Event based churn (teradata / marketing)
Location based ads (EMC / product house)
Social media command center (hadoop / T-SYSTEMS)
Big data fever (2013/14)
Event based churn (teradata / marketing)
Location based ads (EMC / product house)
Social media command center (hadoop / T-SYSTEMS)
DPI reporting (CLOUDERA hadoop / IT)
Big data fever (2013/14)
DPI REPORTING PILOT (2014)
9
DPI REPORTING PILOT (2014)
4bn records per day (1,5 TB raw data)
10
DPI REPORTING PILOT (2014)
Original report: once a year, 3 months delay
4bn records per day (1,5 TB raw data)
11
DPI REPORTING PILOT (2014)
Original report: once a year, 3 months delay
Classic DWH trial: 1 day of data aggregated in 3 days
4bn records per day (1,5 TB raw data)
12
DPI REPORTING PILOT (2014)
Original report: once a year, 3 months delay
Classic DWH trial: 1 day of data aggregated in 3 days
4bn records per day (1,5 TB raw data)
Hadoop trial: 1 day of data aggregated in 15 secs
What is important?
„BIG DAtA analytics is the future of our company”
„FIrst FIX tHE bASICS, BIG DATA and other fancy things can come after that”
What is important?
„BIG DAtA analytics is the future of our company”
„FIrst FIX tHE bASICS, BIG DATA and other fancy things can come after that”
What is important?
„BIG DAtA analytics is the future of our company”
„I have to wait too much for data, i need self service tools”
FUTURE architecture
Self-service framework
Analytical excellence
Data monetization
BI governance
Bingo structure
Future architecture
Data warehouse platform upgrade
SAS environment upgrade
Introduce etl tool
Big data infrastructure
Data quality monitoring
Bi governance
Agile delivery model
Bi governance framework
data governance
Household data model
21
Self-service bi
Business sandbox model
Data discovery tool
Automation framework
Unified Bi portal
governance security privacy
Key challenges
Hadoop in action?
Build data lake
Hadoop in action?
Build data lake Build use cases
Hadoop in action?
Example use case: network customer experience
takeaways
One Data Lake is
essential Develop classic
BI and next generation BI simultaneously
Holistic view of
Big Data is a must for success
Business
should be driving
Thank you
Gergely Vitárius Head of BI Innovation Center [email protected]