What Should a Developer Do With Data?

Poornima [email protected]

@poornima

1

Background

• R&D Engineer @ Synopsys

• Founding Engineer @ Mint.com

• Founder & CEO @ BizeeBee

• Founder of Femgineer.com

• Instructor @ Duke University

2

BIG DATA

3

Not about applications!

4

Agenda

• Not enough data

• Noisy data

• Too much data

• Secured data

5

Process.

No Data

Lots of

Data

Noisy DataSecured Data

Growth Product Launch

Some CustomersTraction

6

No Data or Not Enough?

7

User Experience.

8

Make it compelling.

9

Who’s gonna allow 20-somethings to access their finances?

10

1. Build trust.

11

12

2. Make it frictionless.

13

14

3. Delight!

15

16

17

Privacy.

18

19

Noisy.

Noisy Data

20

Process.

No Data

Lots of

Data

Noisy DataSecured Data

Growth Product Launch

Some CustomersTraction

21

• Data Streams

• Third-Party Data

• User Actions

Noisy Data

22

Parse it. Aggregate it. Mash it up!

23

24

25

Process.

No Data

Lots of

Data

Noisy DataSecured Data

Growth Product Launch

Some CustomersTraction

26

Data.

User data - DB. User data - Analytics.

Application data - Logs.

27

Vocal Minority vs. Major Bug?

28

Mo’ data, mo’ problems!

29

Storage.

30

Retrieval.

31

Warehouse.

32

Distributed computing.

33

Limit the set based on frequency.

34

Process.

No Data

Lots of

Data

Some DataSecured Data

Growth Product Launch

Some CustomersTraction

35

Security - Access Controls.

Employee access. Outsider access.

User access.

36

37

Various Hats

• White Hat

• Black Hat

• Grey Hat

38

Responsible Disclosure.

39

Responsible Disclosure Details

• Driven by social responsibility

• Hardware and software makers repair vulnerabilities

• e.g. Facebook, Google, Mozilla, and Barracuda Networks

40

Review

• Not enough data

• Noisy data

• Too much data

• Secure data

41

Additional Resources

• Office Hours

• Online Mentoring

• Courses

42

Q&A

43