1

Considerations for the Chief Data Scientist

COGNITIVE BUSINESS

Mark WallPartner, IBM Global Business Services

August 2016 @fastbigdata

2

OutlineCognitive Business context Considerations for the Chief Data ScientistCognitive – distinction or continuum from Data Science?How to begin your Cognitive journey

“The productivity slowdown represents a loss of as much as $3 trillion annually in commercial output from 2004 to 2015”.

3

The Solow paradox

4

Facebook Data Scientist salary averages £237,000 pa[link]

The rise of the Chief Data Scientist and a war on talent“We would have more if the talent was there to be had. Last year, the cost of a top, world-class deep learning expert was about the same as a top NFL quarterback prospect. The cost of that talent is pretty remarkable.”

Peter Lee, head Microsoft Research

Analysts IDC predict that by 2018 half of all developer teams will embed “cognitive services” in their apps, up from 1% today (largely based on AI and machine learning technologies) saving $60bn pa by 2020 [link]

5

Views from COO and CEO

Fox, Bob; Ravesh Lala; Owen C. Coelho; Rob van den Dam; and Sandipan Sarkar. “Dialing in a new frequency: Your cognitive future in the communications industry.” IBM Institute for Business Value. December 2015. ibm.com/ business/value/cognitivecommunications

6

Shifting the bell curve of performance

7

Watson Virtual Agent

8

Watson Virtual AgentWith pre-built starter content and ontologies1.Add your own models and data 2.Add external API’s e.g. Computer Vision*, Speech to text, maps 3.Automate your business process and reduce cost to serve

*Computer vision & other Alchemy PhD API’ssee http://blog.alchemyapi.com/the-rise-of-the-phd-apis-new-api-economy

9 Source: IBM Institute for Business Value.

2020 CSP benefits both its customers and an extended ecosystem

10

Design Thinking: People-oriented methods

11

To outthink challenges, competitors and limits, you must conceive of new opportunities you couldn’t imagine before.

12

Reinventing content: Speech to Text, Concept Extraction, Machine Vision at scale

Read more: http://uk.businessinsider.com/ibm-watson-and-ted-talks-2015-5#ixzz3bYHjYiIO*Source: Floridi, Luciano (2014) The Fourth Revolution: How the Infosphere is Reshaping Human Reality

http://watson.ted.com/welcome/

Analysing data from sensors embedded in KONE’s equipment helping to identify and predict issues

Instead of having to send in a service engineer, engineers work with Cognitive analytics to predict and resolve selected technical issues remotely by sensing and deciding on action algorithmically over the Cloud

Reducing downtime, increasing successful outcomes per engineer per day.

Source: Based on a news article originally printed in Fierce Mobile: http://www.fiercemobileit.com/story/ibm-inks-major-watson-iot-deals-finnish-firms/2016-02-19?eid=4437

13

Real example of Cognitive Algorithmics: Next time you step into a lift…

14

Competitors will continue to accelerate disruption and blur the lines among categories

“The biggest threat is new competitors that aren’t yet classified as competitors.”

—Piotr Ruszowski, chief marketing officer,Mondial Assistance, Poland

expect more competitors from outside their industry, while only 29 percent expect more competition from within their industry.

54%of CxOs

SOURCE cited in notes

15

Three- to five-year strategiesfor CxOs:

They’ll be hampered by limited insights when only:

CxOs sense the opportunity, but are limited by lack of visibility

80 percent are set on being the first to market with innovation

66 percent plan to focus more on customers as individuals

81 percent expect to shift to more digital, virtual client engagement models

51 percent draw on customer feedback

39 percent draw from adjacent industries

29 percent draw from blogs and social media sites

SOURCES cited in notes

+

16

+

Your digital intelligence is your competitive advantage

Structured and active Unstructured and dark

17

Think of all that has been accomplished using only

a fraction of the available data

Unlock the possibilities.

Only cognitive unlocks the potential in all data

What answers lie in the

88% that is dark?1

By the year 2020, about 1.7 MB of new information will be created every second, for every human being on the planet.2

SOURCES cited in notes

18

Cognitive systems are fundamentally different from what you have today

Adapt and make sense of all data; “read” text, “see” images and “hear” natural speech with context

Understand

Reason Interpret information, organize it and offer explanations of what it means, with rationale for the conclusions

Learn Accumulate data and derive insight at every interaction, perpetually

19

Add completely new abilities to your business technologies

Programmable computing responds to requests and makes determinations, analyzing data according to predefined parameters.

Cognitive systems

interact with humans naturally to interpret data,

learning from virtually every interaction and proposing new possibilities through

probabilistic reasoning.

Shift technology’s role

from enabler to advisor

Cognitive business The next evolution of human

and systems capabilities, where technology enhances, scales and

accelerates human expertise

20

21

Enhance, scale and accelerate human expertise to empower all people, all roles.

Go beyond analysis to hypothesis, conclusion and action—in weeks instead of months or years

Collate decades of knowledge and data to create an evidence-based, virtual advisor and elevate entire

teams to the level of your best experts

Teach your system to learn and uncover patterns and insights from all kinds of information such as research data, images and notes

Build workflows that can be coached by humans to grow ever more effective, safe or productive with each interaction

22

INTERACT

SENSE

Ontologies Knowledge graphsConcept Extraction

Semantic Web

DECIDE

Sour

ce: A

dapt

ed fr

om C

ollin

s, R

. (20

14) G

over

nanc

e in

the

age

of B

ig D

ata

–A P

oint

of V

iew

(IB

M G

BS

)

UNDERSTAND & INFER

Role of Cognitive

Learning from interaction

Manage Cognitive

Considerations for the CDO: The role of analytics, BI and cognitive in performance management UNDERSTAND &

DECIDE

23

The role of analytics, information in Cognitive Business• Analytics objective: “Better” decision making leading to better business outcomes• Cognitive is different

• scales the organization aptitude for total and instant recall of all available knowledge • Supports real time decision making and learning across the Enterprise in discursive interactive

manner: Natural Language Processing + Machine Learning• Influence decisions directly inside the process or longer term exploration and discovery to allow the

improvement of processes• Cognitive capability is a learning capability.• Analytics is Determinative in decision making (prescriptive analytics) • Cognitive tends to influencing “soft” human expert decision processes either individual or

collective.• Quality of the decision depends on the quality of the data and information used and the power of the

analytic test or technique used• Not only do business need to understand how information and analytics support decision making and

risk taking but increasingly need to justify this to regulators and others

24

Distinction of Analytics+BI and Cognitive Analytics & BI Cognitive

Data Structured, 4’Vs, metadata, integration, data movement

Corpus, concept extraction and understanding, semantics;RDF’s Cognitive API’s & Algorithmics “Just add data”, interoperability of processes through data

Models ETL, Normalisation, batch, procedural

Unstructured, analytics as the model, Machine learning and Natural Language Processing, on-demand, linked data, federated, information inference

Answers Deterministic, rules driven, data driven

Probabilistic, data ‘aware’

Impact on Governance

roles

Stewardship of Data processes (lineage)

Additional Stewardship of Data Provenance; ability to explain how the answer was derived on demand at a particular point in time;

Architectures ETL, EDW, 3NF Ontologies, Data Fabric

Services Reporting, Service Oriented architecture (SOA)

Decomplet Micro services; Service Enabled Architecture; Resource Oriented Architecture (ROA) & RDF’s

25

…Predictive analytics weaved together with Cognitive analytics to personalize and enhance the experience

Customer Segmentation

ActionClustering

Personality Analytics

Sentiment Analytics

ImageRecognition

Next BestAction

Predictive Analytics Cognitive Analytics

Internal Data Syndicated Data Social Data Device Data

26

Today’s industry leaders recognize the gap in their organizations’ abilities

50%say available data limits confidence in strategic decisions

95% plan to invest in cognitive

Healthcare

SOURCES cited in notes

believe they can’t deliver on consumer expectations

94% plan to invest in cognitive

Retail

60% 30%say the quality of data is insufficient for business model innovation

98% plan to invest in cognitive

Insurance

27

Seize opportunities for disruption before your competitors do

Amplify knowledge,reimagine workflows

Transform the enterprise

Cogn

itiv

e Bu

sine

ss

Accelerate learning andscale expertise

Institute processes and operations with learningbuilt in

Deepen relationships through interactive, personal engagement

Develop applications, products and services that read, see, talk, hear and learn

Discover and explore intelligently to unlock new business models and accelerate evolution

Become indispensable to users and customers

Give workflows cognitive capabilities

DataIBM Watson Developer Cloud IBM Bluemix®

IBM Watson Ecosystem

Build cognitive applications today with Watson APIs

28

Data

IBM Cognitive Business Solutions

Watson Engagement Advisor

Watson Health

Watson Discovery Advisor

IBM Power Systems™ and Storage

IBM Global Technology Services

Construct a cognitive system

Activate cognitive technology in your business with the right ingredients

Hybrid Cloud Infrastructure (SoftLayer®, IBM Systems)

Security Intelligence (QRadar®)

DataIBM Marketing and Commerce clouds IBM advanced analytics solutions

Partner apps, Powered by Watson

IBM Internet of Things Foundation

Watson Explorer

IBM Global Business Services

IBM Global Technology Services

Outthink boundaries and unleash new opportunities with IBM

29

IBM Watson portfolio

IBM strategic and data partnerships

IBM Watson Ecosystem

IBM Watson Developer Cloud

IBM cognitive expertise and research

The cognitive era is here. You must be an active participant in it to be a beneficiary of it.

@FastBigData 30

Welcome to the cognitive era

Join us on the journey

31

Trademarks and notes

IBM Corporation 2016

• IBM, the IBM logo, ibm.com, and Watson are trademarks or registered trademarks of International Business Machines Corporation in the United States, other countries, or both. If these and other IBM trademarked terms are marked on their first occurrence in this information with the appropriate symbol (® or ™), these symbols indicate U.S. registered or common law trademarks owned by IBM at the time this information was published. Such trademarks may also be registered or common law trademarks in other countries. A current list of IBM trademarks is available on the Web at “Copyright and trademark information” at www.ibm.com/legal/copytrade.shtml.

• Other company, product and service names may be trademarks or service marks of others.

• References in this publication to IBM products or services do not imply that IBM intends to make them available in all countries in which IBM operates.

Appendix

32

Pick your starting point

Evaluate and curate data and knowledge

Ready your organization

Apply cognitive technology

Enhance and expand

“I want to become indispensable to users and customers.”Build cognitive apps today with Watson APIs

Select the appropriate Watson APIs to build your cognitive app; identify the knowledge and content to form a knowledge base - collect both internal and external data sources.

Configure and train new cognitive applications.

Test and validate your new cognitive app through end user engagement.

Validate usage benefits. Deploy and expand to other domains across the organization.

“I want to amplify knowledge and reimagine workflows.”Give workflows cognitive capabilities

Locate and curate data; identify gaps and prepare for integration with existing systems.

Define the community and implementation plan – emphasizing the newvalue from cognitive workflows.

Deploy and manage, while driving usage and iterative learning.

Decide if the cognitive workflow you’ve deployed creates new opportunities for further development and adoption.

“I want to transform the enterprise.”Construct a cognitive system

Complete formal strategy assessment. Collect, ingest, curate, annotate and build out enterprise-wide taxonomies and ontologies.

Adopt and invent all-new processes, content and roles that will enable you to think and act differently in a cognitive era.

Define and execute staged roll-out. Instrument metrics and KPIs.

Collect and address metrics and KPIs. Periodically update functionality and training to tune for new content. When ready, deploy and expand to other domains.

Assess your foundation Data Development Hybrid cloud

infrastructure Security Education

Develop your cognitive strategy

Identify a problem to solve that will lead to critical breakthrough

opportunities.

Cast a vision. State a clear use case, secure

executive support, plan for adoption and

deployment.

Champion a new culture. Prepare

people for new ways of collaborating with

technology.

How to begin your cognitive journey

Let cognitive abilities take root to set disruption in motion.

34

“IBM’s Watson is already being used to apply cognitive computing in a wide range of industries and technologies, from education to banking to winemaking to urban planning, and so much more.” Fast Company

“Cognitive everything—by 2018, over 50 percent of developer teams will embed cognitive services in their apps (versus 1 percent today), providing U.S. enterprises with over USD$60 billion in annual savings in 2020.” IDC

“We wanted it to be more personal and intuitive with natural language…. the results change dramatically as more questions are asked. And customers can ask questions which would not be possible with filters.” Cal Bouchard, The North Face

SOURCES cited in notes