mike olson, cloudera

18
Big Data, Bigger Ques.ons CDO Leader Forum | February 10, 2015 Mike Olson | Chief Strategy Officer, Cloudera

Upload: corinium-global

Post on 14-Apr-2017

61 views

Category:

Data & Analytics


0 download

TRANSCRIPT

Page 1: Mike olson, cloudera

Big  Data,  Bigger  Ques.ons  CDO  Leader  Forum  |  February  10,  2015  

Mike  Olson  |  Chief  Strategy  Officer,  Cloudera  

Page 2: Mike olson, cloudera

2  ©  Cloudera,  Inc.  All  rights  reserved.  

Data  can  be  a  powerful  strategic  asset  

data  helps  achieve  your  business  vision.  

…only  if...  

Page 3: Mike olson, cloudera

3  ©  Cloudera,  Inc.  All  rights  reserved.  

Data  Changes  How  We  Work  

Everything  that  can  be  measured  will  be  measured.  

Employees  and  customers  expect  more  personal  interac.ons,  but  not  at  the  cost  of  their  privacy.  

The  most  innova.ve  companies  embrace  experimenta.on  and  agility.  

Instrumenta.on   Consumeriza.on   Experimenta.on  

Page 4: Mike olson, cloudera

4  ©  Cloudera,  Inc.  All  rights  reserved.  

Data  Sources  

Data  Systems  

Data  Access  

Business  Analy.cs  

Custom  Applica.ons  

Exis.ng  Data  

Databases  

Opera.onal  Applica.ons  

New  Data  

Limited  Data  Not  efficient  to  keep  exis.ng  data,  let  alone  handle  new  data  sources.  

Time  consuming  to  transform  data  for  analysis  in  exis.ng  systems.  

Limited  Insights  Power  users  struggle  with  data.  

Many  users  have  no  data.  

Compliance  and  Privacy  More  data,  more  users,  and  more  tools  create  complexity.  

Need  to  balance  business  agility  with  security  and  governance.  

Tradi.onal  Architectures  Under  Pressure  

Page 5: Mike olson, cloudera

5  ©  Cloudera,  Inc.  All  rights  reserved.  

A  New  Architecture:  The  Enterprise  Data  Hub  

A  new  kind  of  data  plaYorm.  • One  place  for  unlimited  data  

• Unified,  mul.-­‐framework  data  access  

Enterprise-­‐Grade:  •  Leading  performance  •  Compliance-­‐ready  administra.on  and  data  management  

•  Fundamentally  secure  

•  Open  source,  open  standards  

Security  and  Administra.on  

Unlimited  Storage  

Process   Discover   Model   Serve  

Deployment  Flexibility  

On-­‐Premises  Appliances  Engineered  Systems  

Public  Cloud  Private  Cloud  Hybrid  Cloud  

Page 6: Mike olson, cloudera

6  ©  Cloudera,  Inc.  All  rights  reserved.  

The  Importance  of  Being  Mul.-­‐Framework  

Batch  Processing  

Interac.ve  SQL  

Search  

NoSQL  

Stream  Processing  

Machine  Learning  

Highly  mature  for  loading  and  processing  large  amounts  of  data  

Self-­‐service  BI  for  analysts  to  quickly  explore  and  analyze  data  

User-­‐friendly  search  for  business  users  to  quickly  access  data  

Real-­‐.me  single  event  querying  at  high  volumes  

Robust,  real-­‐.me  querying  on  collec.ons  of  events  

Quick  model  itera.on  for  data  scien.sts  for  advanced  analy.cs  

Hourly  repor.ng  

Near  real-­‐.me  BI  

Cross-­‐applica.on  search  

Real-­‐.me  pa`ern  recogni.on  

Predic.ve  analy.cs  

Advanced  model  building  

Page 7: Mike olson, cloudera

7  ©  Cloudera,  Inc.  All  rights  reserved.  

Comprehensive,  Compliance-­‐Ready  Security  Authen.ca.on,  Authoriza.on,  Audit,  and  Compliance  

Perimeter  Guarding  access  to  the  cluster  itself  

Technical  Concepts:  Authen.ca.on  

Network  isola.on  

Access  Defining  what  users  and  applica.ons  can  

do  with  data  

Technical  Concepts:  Permissions  Authoriza.on  

Data  Protec.ng  data  in  the  

cluster  from  unauthorized  visibility  

Technical  Concepts:  Encryp.on,  Tokeniza.on,  

Data  masking  

Visibility  Repor.ng  on  where  data  came  from  and  how  it’s  being  used  

Technical  Concepts:  Audi.ng  Lineage  

Cloudera  Manager   Apache  Sentry   Cloudera  Navigator  Navigator  Encrypt  &  Key  

Trustee  |  Partners  

Page 8: Mike olson, cloudera

8  ©  Cloudera,  Inc.  All  rights  reserved.  

Data  Sources  

Data  Systems  

Data  Access  

Business  Analy.cs  

Custom  Applica.ons  

Exis.ng  Data  

Databases  

Opera.onal  Applica.ons  

New  Data  

Keep  Unlimited  Data  From  disparate  and  limited  views,  

to  unlimited  informa.on  access.  

Unlock  Value  from  Data  From  analy.cs  for  some,  

to  insights  for  all.  

Manage  Compliance  From  risk  due  to  regula.ons  and  customer  privacy  concerns,  

to  trust  in  a  secure  and  compliant  plaYorm.  

Enterprise  Data  Hub  

Security  and  Administra.on  

Unlimited  Storage  

Process   Discover   Model   Serve  

More  Value  from  More  Data  for  More  People,  Faster  

Page 9: Mike olson, cloudera

9  ©  Cloudera,  Inc.  All  rights  reserved.  

The  Value  of  an  Analy.cs  Strategy  

Build  data  value  for  customers  and  employees.    

Remove  uncertainty  from  the  business.    

The  most  valuable  companies  embrace  experimenta.on  and  agility.  

Increase  Revenue   Decrease  Risk   Accelerate  Innova.on  

Page 10: Mike olson, cloudera

10  ©  Cloudera,  Inc.  All  rights  reserved.  

Automated  analy.cs  at  users’  finger.ps  

What  SHOULD  happen  

What  IS  happening  

What  DID  happen  

$500M  in  averted  energy  spend  

What  WILL  happen  

CiEzen  

Page 11: Mike olson, cloudera

11  ©  Cloudera,  Inc.  All  rights  reserved.  

The  Pervasive  Analy.cs  Journey    

Page 12: Mike olson, cloudera

12  ©  Cloudera,  Inc.  All  rights  reserved.  12  

How  do  seed  selec.on,  plan.ng  density,  irriga.on,  ground  temperature,  soil  chemistry  and  weather  impact  yields?  

How  much  corn  did  my  farm  produce  last  year?  

Sample  fields  at  fine  resolu.on  and  design  a  plan.ng  strategy  to  increase  yields  while  conserving  water  and  chemicals.  

Page 13: Mike olson, cloudera

13  ©  Cloudera,  Inc.  All  rights  reserved.  13  

How  to  demographics,  lifestyle,  medical  history  and  environmental  factors  impact  heart  disease  in  pa.ents  like  this  one?  

Do  this  pa.ent’s  symptoms  indicate  heart  disease?  

Use  personal  monitoring  devices  and  social  media  to  track  the  pa.ent’s  condi.on  and  manage  chronic  disease  to  be`er  outcomes.  

Page 14: Mike olson, cloudera

14  ©  Cloudera,  Inc.  All  rights  reserved.  

What  can  we  learn  from  using  much  larger  and  more  varied    data  sets  for  advanced  security  and  threat  analy.cs?  

How  much  can  we  cut  our  storage  footprint  and  costs  if  we    increase  governance  of  ac.ve  data  rather  than  archive?  

Curtail  $30  million  fraud  case  –  largest  in  company  history.  Create  $1  billion  data  product  offering  –  not  previously  possible.  

Page 15: Mike olson, cloudera

15  ©  Cloudera,  Inc.  All  rights  reserved.  

Can  we  capture  more  detailed  data  streams  to  personalize    policies  to  actual  day-­‐to-­‐day  occurrences  at  each  property?  

How  do  we  use  standard  profile  informa.on  for  each    house  to  determine  risk  and  set  individualized  rates?  

Scale  to  run  models  across  data  from  all  50  states  simultaneously.  Experience  an  average  7500%  speed-­‐up  on  descrip.ve  analy.cs.  

15  © 2014 Cloudera, Inc. All rights reserved.

Page 16: Mike olson, cloudera

16  ©  Cloudera,  Inc.  All  rights  reserved.  

What  opportuni.es  are  we  not  seeing?  Can  we  iden.fy  and    inves.gate  anonymous  pa`erns  or  trends  in  real  .me?  

Can  we  eliminate  sampling  error  by    including  all  our  log  data  in  analyses?  

Iden.fy  and  isolate  high-­‐value  pa`erns  without  pre-­‐assignment.  Build  real-­‐.me  recommender  systems  to  op.mize  buy/sell.  

16  © 2014 Cloudera, Inc. All rights reserved.

Page 17: Mike olson, cloudera

17  ©  Cloudera,  Inc.  All  rights  reserved.  

Pa`erns  &  Predic.ons  –  Full  Bleed  

Can  we  use  real-­‐.me  predic.ve  modeling  and  machine  learning  to  iden.fy  cri.cal  correla.ons  between  veterans’  

communica.ons  and  mental  health?  

Page 18: Mike olson, cloudera

Thank  You!  Mike  Olson,  Chief  Strategy  Officer  

[email protected]  @mikeolson