dataiku, pitch at data-driven nyc, new york city, september 17th 2013

15
DATA SCIENTIST is NOT a defined term This is not… a Data Scien9st www.dataiku.com @dataiku @baAymarc

Upload: dataiku

Post on 18-Dec-2014

716 views

Category:

Technology


0 download

DESCRIPTION

Our pitch at Data-Driven NYC meetup on September 17th (http://datadrivennyc.com). Speaking about Data Scientists pains and how Dataiku Data Science Studio can help them to more than Data Cleaners and Data Leak Fixers !

TRANSCRIPT

Page 1: Dataiku, Pitch at Data-Driven NYC, New York City, September 17th 2013

DATA  SCIENTIST  

is    NOT  

a  defined  term   This  is  not…  a  Data  Scien9st    

www.dataiku.com    -­‐  @dataiku  -­‐  @baAymarc  

Page 2: Dataiku, Pitch at Data-Driven NYC, New York City, September 17th 2013

MACHINE  LEARNING  EXPERT  

Page 3: Dataiku, Pitch at Data-Driven NYC, New York City, September 17th 2013

DATA  CLEANER  

Page 4: Dataiku, Pitch at Data-Driven NYC, New York City, September 17th 2013

DATA  LEAK  FIXER  

Page 5: Dataiku, Pitch at Data-Driven NYC, New York City, September 17th 2013

END  OF    (HADOOP)  JOB    

DATA  WAITER  

Page 6: Dataiku, Pitch at Data-Driven NYC, New York City, September 17th 2013

How  can  we    HELP    

DATA  SCIENTISTS  to    

FOCUS  on  the    

REAL  PROBLEMS  ?    

Page 7: Dataiku, Pitch at Data-Driven NYC, New York City, September 17th 2013

Pain  points  

•  Data  prepara9on  is  9me-­‐consuming    •  Machine  learning  is  hard  to  understand  

•  Insights  and  models  (almost)  never  reach  produc9on  

Page 8: Dataiku, Pitch at Data-Driven NYC, New York City, September 17th 2013

Data  Science  Studio  

•  A  democra9c  &  ready  to  use  Data  Science  Studio  to  start  innova9ng  with  data!  

Ready  to  Use  Data  Science  PlaYorm  

Common  playground  for  innova9on  

Accessible  Sta9s9cs  &  Machine  Learning  for  

everyone  

Handle  real-­‐life  data  

Page 9: Dataiku, Pitch at Data-Driven NYC, New York City, September 17th 2013

Data  Science  Studio  Visual  and  Interac9ve  Data  Prepara9on  For  Data  Cleaners  

Guided  Machine  Learning  For  non  Machine  Learning  Experts  

Produc9on  ready  For  Data  Leak  Fixers  

Page 10: Dataiku, Pitch at Data-Driven NYC, New York City, September 17th 2013

Visual  Data  Prepara9on  

Page 11: Dataiku, Pitch at Data-Driven NYC, New York City, September 17th 2013

Visual  Data  Prepara9on  

•  Interac9ve  UI  with  instant  feedback  and  sugges9ons  

•  Reversibility  of  the  script,  data  integrity  •  Explora9on  of  data:  quick  analysis,  facets  •  Cleansing:  missing  values,  outliers,  parsing  •  Enrichment:  GeoIP,  Holidays,  joins  •  Produc9on-­‐ready:  integra9on  within  a  flow  

Page 12: Dataiku, Pitch at Data-Driven NYC, New York City, September 17th 2013

Guided  Machine    Learning  

Page 13: Dataiku, Pitch at Data-Driven NYC, New York City, September 17th 2013

Produc9on  &  orchestra9on  

Page 14: Dataiku, Pitch at Data-Driven NYC, New York City, September 17th 2013

Data  Science  Studio:    benefits  •  Real-­‐9me  and  interac9ve  

–  Transforma9on  effects  can  be  previsualized  in  real-­‐9me    •  Transparent  and  traceable  

–  Keep  the  full  history  of  your  data  transforma9on  logics  and  model  designs  

•  Easy  access  to  machine  learning  –  Get  started  with  our  app  templates,  bootstrap  your  model  and  features  selec9ons,  then  go  further!  

 •  Scalable  and  Produc9on  Ready  

–  Apply  your  recipes  on  your  cluster  on  terabytes  of  data  

Page 15: Dataiku, Pitch at Data-Driven NYC, New York City, September 17th 2013

Dataiku  at  a  glance  •  Founded  in  2013  by  Data  and  Search  Engine  veterans  •  From  “data”  and  “haïku”  

“data  can  be  big    solu;on  would  be  small  

feel  the  hot  wind”  

•  1  goal:  make  Data  Science  accessible  to  anyone!              

Contact:  [email protected]  -­‐  @baAymarc  -­‐  github.com/dataiku