Based on as my experience, I am providing high level quick presentation that introduces how IBM has moved from Jeopardy and DEEP Blue to era of Cloud and Microservices and AI.
1. Cognitive & AI & Cloud
Overview
Prague, June, 2018
Martin Svik, PhD. MBA.
Exec IT Architect, WW Strategy
Watson iLAB Leader
2. 1. Watson Cognitive Technologies Overview
2. Virtual Assistants/Chatbots
3. Why IBM
4. Typical Use Cases, Examples and Case Studies
Content
3. REASONUNDERSTAND LEARN INTERACT
Cognitive represents a new era of Systems that …
… and create a new partnership between humans and machines
HUMANS
MACHINES
+
HUMANS
MACHINES
OR>
5. Cognitive is a set of technologies and approaches, Watson is an
umbrella
How Watson sees, hears, speaks, feels, translates, finds …
6. Experts
that understand your work and offer
answers and insights from Watson’s
ecosystem of sources and expertise
Personal Assistants
that will help prioritize work and handle
common tasks like scheduling meetings
and returning emails, surface functions
based on behavior
IBM provides two essential, new, cognitive team members
7. Watson
Conversation
Here Watson uses
reasoning strategies
that focus on the
language and
context of the
question.
Watson
Discovery
Service
Here Watson uses
reasoning strategies
that focus on
identifying the most
appropriate answer.
Basic & Enhanced Conversational Intelligence
• Some intents are repeated over and over and can be modeled
• But there is a “long tail” of infrequent intents. Modeling the long tail is impossible and requires a different approach
Unique Intents
100s 100,000+
Volume,
Frequencyof
Questions
Long Tail
Turn on my headlights.
My exhaust is making a rattling sound, how do I
troubleshoot the problem?
Short Tail
Majority of
questions hit a
small number of
intents
Different Approach required to
bring back appropriate
response given number of
answers
8. Conversational Computing Platforms, Q2 2018
IBM Leads The Pack
Forrester’s research uncovered a market in which IBM leads the pack. Amazon,
Google, Microsoft, and Nuance Communications offer competitive options. Oracle
offers a strong option for certain applications, while Rulai lags.
The Market Is Still In Its Infancy
As this nascent market matures, some vendors will specialize, and others will flex
their muscles. Amazon and Google, in particular, are poised to exploit their
dominant market positions in cloud, data, and AI.
9. HfS Blueprint Report - Enterprise AI Services 2018
Leading global service provider driving AI at scale with differentiating
assets around Watson ecosystem.
Strong association with cognitive and AI
Deep investments in Watson capabilities
Formalized AI practice
Driving scale through focus on core technologies
Broad set of vertical and domain-specific offerings
10. Typical Use Cases
Virtual customer sales advisor
§ Answer questions about products, services, etc.
§ Answer questions based on personal information and situation/history.
§ Support execution of end-to-end tasks on behalf of the client.
§ Guide customers during subscription processes.
§ Recommend next best actions to the customers.
§ Reduce churn of existing customers
§ Increase customer lifetime value through
upsell and cross-sell
§ Create new revenue streams by
understanding your clients’ needs better
§ Reduce cost to serve
Virtual customer helpdesk
§ Answer repetitive questions.
§ Answer questions based on personal information and situation/history.
§ Support execution of end-to-end service requests.
§ Capture and solve clients complaints, or handover to agent.
§ Increased first contact resolution with
contextually relevant answers
§ Improved Customer satisfaction
§ Reduction in cost for operations
Enable employees with
knowledge
§ Provide actionable and relevant insights (e.g. Virtual Assistance for
Relationship Managers in banking and insurance)
§ Analyse unstructured data, extract insights and evidences based
recommendations (e.g. Analysis of Credit Contracts)
§ Leverage the best available expertise and knowledge
§ Speed and quality of decision making and
advice
§ Lower cost of information gathering and
access
§ Reducing research time for Advisors
Internal helpdesk for employees
§ Answer frequent questions.
§ Troubleshoot problems based on information about employees assets.
§ Support execution of end-to-end service requests.
§ Capture employees requests.
§ Improves employee and business
productivity
§ Reduces Level 1 helpdesk costs
§ Provides full time virtual helpdesk availability
§ Increases productivity, allowing SMEs to
concentrate on value-add activities
11. Martin Svik, PhD. MBA.
Executive IT Architect, WW strategy
email: martin.svik@cz.ibm.com
mobile: +420737264389
LN: https://www.linkedin.com/in/martinsvik
Thank you