NCUR Presentation Walkmore

A REPORT FROM THESINGULARITY

SU?

TO  USE  EXPONENTIALLY  GROWING  TECHNOLOGIES  TOPOSITIVELY  IMPACT  1  BILLION  PEOPLE  IN  TEN  YEARS

80  STUDENTS  FROM  36  COUNTRIES  BROUGHT  TOGETHER

“EXPONENTIALLY  GROWING  TECHNOLOGIES?”

COMPUTERS  DOUBLE  THEIR  PROCESSING  POWER  EVERY  TWO  YEARS.

“There  is  a  world  market  for  maybe  five  computers.”

-­‐  Tom  Watson,  Founder  of  IBM

“Cinema  is  a  fad.”-­‐  Charlie  Chaplin

“Flight  by  machines  heavier  than  air  is  unpracZcal  and  insignificant,  if  not  

u[erly  impossible.”  -­‐  Canadian  mathemaZcian  Simon  Newcomb,  18  months  before  the  Wright  

Brothers’  flight

SINGULARITY TRAINS BUSINESS LEADERS, ENTREPRENEURS, SCIENTISTS, &

INNOVATORS TO THINK EXPONENTIALLY.

MOST BUSINESSES, GOVERNMENTS, SYSTEMS PLANNERS THINK LINEARLY

ABOUT THE FUTURE OF TECHNOLOGY.

Dr. Philipp KallerhoffJumiya, Inc

Eli MohamadJumiya, Inc

MEET  CHRIS

THE  INTERSECTION  OF  SYSTEMS

FOR  EXAMPLE:CHRIS’  DOCTOR  BILLS  WERE  SENT  STRAIGHT  TO  A  COLLECTION  AGENCY,  HURTING  HIS  CREDIT  BY  100  

POINTS.

62OF  BANKRUPTCIES  ARE

BECAUSE  OF  MEDICAL  DEBTAMERICAN JOURNAL OF MEDICINE, 2012

%

CREDIT  SCORING  IS  A  BLACK  BOX

Meet the gatekeepers.

IRMA  WORKED  WITH  CHRIS  AND  WALKED  WITH  HIM  

HAND  IN  HAND.

HER  WORK  REQUIRES  AN  UNDERSTANDING  OF  HOW  MULTIPLE  SYSTEMS  INTERSECT  AND  EFFECT  

CHRIS’  LIFE.

BUT  IRMA’S  ONLY  ONE  PERSON.

HOW  CAN  WE  BETTER  UNDERSTAND  HOW  COMPLEX  SYSTEMS  AND  BEHAVIORS  

IMPACT  OUR  LIVES?

MACHINE  LEARNING  ISFINDING  PATTERNSIN  COMPLEXITY

NETFLIX & AMAZON RECOMMENDATIONS

WITH  PEOPLE  LIKE  CHRIS  IN  MIND:  OUR  TEAM  USED  MACHINE  LEARNING  EXPLORED  THE  CONNECTIONS  IN  HEALTH  AND  FINANCE.

SURVEYED  HEALTH

SURVEYED  FINANCE

$

THE  TEAM  FOUND  PATTERNS

RUNNING  HAS  6x  CORRELATION  WITH  TIMELY  BILL  PAYMENT

DRINKING  HAS  A  6x  CORRELATION  WITH  BANKRUPTCY

$80%

ACCURACY

HOW  CAN  WE  LEARN  MORE  ABOUT  THESE  CONNECTIONS?

FIRST:

250  million  transac5ons500,000  users  

Can  we  predict  member  behavior,  improve  profitability,  and  reduce  risk  by  looking  at  a  dataset  including:

PREDICT  DEFAULTS  WITH  80%  ACCURACY

RESULTS  DIFFERED  FROM  TRADITIONAL  CREDIT  SCORE  BY  10%

MODEL  APPROVES  1%  MORE  LOANS  AND  REDUCES  DEFAULTS  BY  15%

“For  example,  5%  of  your  members  would  be  10%  more  profitable  if  they  used  a  loan  product.  15%  of  

those  members  are  very  likely  to  sign  up  for  the  product,  and  using  our  new  credit  risk  model  we  can  approve  10%  

more  people  for  the  loan”

PREDICT  PRODUCT  LIFE  CYCLE  •  IMPROVE  PROFITABILITY  •  REDUCE  RISK

For  example:

THE  FIRST  MINIMUM  VIABLE  PRODUCT

NEXT:CONNECT  THESE  FINDINGS  WITH  

WELLNESS  BEHAVIOR.

The  account  that  improves  as  you  do.

Enable  members  to  increase  and  improve  access  to  financial  products  and  be  rewarded  for  a  healthier  lifestyle.  

42

By  engaging  in  healthy  behavior,  members  can  posi5vely  impact  their  financial  access  by  

improving  rates  and  balances.

43

44

Use  data  from  devices  like  Nike+,  Fitbit,  Runkeeper,  or  100+  other  providers...

...or  just  your  phone...

Purpose

1. Generate  new  revenue  from  exisZng  products2. Help  retain  exisZng  and  a]ract  new  CU  members  3. Improve  overall  member  profitability,    reduce  lending  risk,  and  increase  repayment

Calls  to  ac5on

• The  research  suggests  healthy  living  members  are  be]er  risk  and  more  profitable• Researchers  were  able  to  predict  the  next  best  product  using  transacZonal  data• The  research  suggests  that  quanZfied  self  data  can  improve  the  predic5on  model  even  further• AddiZonal  data  sets  -­‐  economic  or  locaZon  data,  for  example  -­‐  impact  on  consumer  behavior.

52

IS  THERE  A  MEASURABLE  CONNECTION  

BETWEEN  PHYSICAL  AND  FINANCIAL  

HEALTH?

YESCAN  WE  PREDICT  

BEHAVIOR  BASED  ON  THOSE  

CONNECTIONS?

YESCAN  WE  USE  DATA  TO  PREDICT  AND  

IMPROVE  MEMBERS  FINANCIAL  HEALTH?

YES

 CAN  WE  INTEGRATE  HEALTH  AND  FINANCIAL  DATA  TO  MOVE  MEMBERS  ALONG  A  PATH  TO  TOTAL  WELLNESS?

WE  NEED  OPT-­‐IN  MEMBER  DATA

JOIN  THE  PILOT  PROJECT

Calls  to  ac5on

Get  involved  in  the  pilot  project• Pilot  starts  in  July  and  will  last  9  months• 3  credit  unions,  with  1  more  spot  available• All  necessary  infrastructure  provided  through  plagorm  at  www.walkmo.re• The  plagorm  will  be  adapted  to  fit  each  credit  union’s  needs

h]p://[email protected]

[email protected]

[email protected]

QUESTIONS?  THANK  YOU!