9
K L Mukesh

Business Intelligence on Cloud: A Business Perspective

Embed Size (px)

DESCRIPTION

TDWI India Chapter 2011 Feb 05 Hosted at Intel, Presentation from K L Mukesh, Tata SKy

Citation preview

Page 1: Business Intelligence on Cloud: A Business Perspective

K L Mukesh

Page 2: Business Intelligence on Cloud: A Business Perspective

Young company of 5 years; Started with basic performance reports like sales &

installations Few months down the line started dealer performance

reporting Started measuring Call center performance Started with ROI reporting Supposed to generate 25-30 reports; Number of reports

and dashboard growing by the day

ETL challenges – technically feasible, cost and time an issue

Page 3: Business Intelligence on Cloud: A Business Perspective

Data

Reports Information

Knowledge

InsightsActions

Page 4: Business Intelligence on Cloud: A Business Perspective

Selected one of the vendors for BI, considerations were

• Speed to Market – Few weeks, rather than many months• Lack of internal expertise in BI• Drastic fluctuations in end user requirements • Disparate set of metadata within the enterprise• Predictive Modeling

Page 5: Business Intelligence on Cloud: A Business Perspective

Upside: Data Transfer to cloud (Vendor location) streamlined

in weeks Started with Customer Analytics Moved to Predictive modeling End to end integration – analytics to call center for

actions back to analytics for performance tuning ARPU significantly above industry average

Downside: Scaling horizontally across functions a challenge;

Page 6: Business Intelligence on Cloud: A Business Perspective

Small applications with highly distributed usersmoving to the cloud – dealer services, managingfield services

Data mart / performance dash boards associatedwith these applications a natural fit

BI on cloud will follow transaction based systemson cloud; Easy BI access to distributed users;

Enterprise view a challenge. Hopefullyarchitecture will evolve over time

Page 7: Business Intelligence on Cloud: A Business Perspective

Large data set options where time lagacceptable being explored.◦ Customer profiling◦ Churn prediction◦ Campaign performance◦ Processing capacity intensive – statistical modeling

Loading data once in a fortnight acceptable.Significant capacity requirement variation

Data moved using physical medium

Page 8: Business Intelligence on Cloud: A Business Perspective

Yes No

Yes Yes

APPLICATION SIZE

SMALL LARGETURN

AROUND

TIME

REALTIME

TAKEYOURTIME

Page 9: Business Intelligence on Cloud: A Business Perspective

Thank You