Watson DevCon 2016 - Improve User Engagement with Watson API Mashups
Presentation given by Giuliano Iacobelli at Watson DevCon 2016.
Building a chat bot that can hear and talk — easy. Building a chat bot that can listen and respond sympathetically — not as easy. Until now. In this session, you'll discover how Watson APIs can be combined with Watson Conversation to create chat bots that lend a human touch to automated interactions. Learn the anatomy of a conversational app, understand where AI APIs come into play, and see how services like Watson Tone Analyzer, which analyzes linguistic tones in text, can create chat bots that engage users on a deeper level.
More than 900 million monthly active
users only on Facebook Messenger
In 2015 messaging apps have been the
fastest growing In the global Top 10 of
most used apps 6 are messaging apps
No need to download yet another app
2016 is the year
of the bots
Messaging platforms aim to change the way
brands communicate with consumers
Chat UX is more natural, dynamic and avoid
any friction opening new accounts
Do more from a familiar environment
where we already spend our time.
2016 is the year
of the bot
Where you receive messages and route them to an appropriate handler.
The point of the Messenger is to act as an interfact on top of our bot.
Message Parser & Handler
The role of the handler is to handle the message passed from Messenger,
delegate tasks to services, and send messages back to the user.
In charge of making API requests, parsing data, and formatting responses for our handlers.
The line between the responsibilities of message handling and actions.
Stores help manage state from interactions as your bot might need to recall things. Stores
hold conversation data and keep track of what state the system is in.
Chatbots are entirely API-driven & event-driven
If you’re the type of developer who hates ﬁddling with CSS,
this might be the happiest day of your life.
Receive message from user
Update conversation state
Reply to user
to third party system
Understand request with Watson
“The times are difﬁcult! Our sales have been disappointing for the past three
quarters for our data analytics product suite. But we are not doing a good job
at selling it and this is really frustrating.”
IBM Watson Tone Analyzer
Analyzes text at the document level and the sentence level to get a sense of the overall tone of
the document and identify speciﬁc areas of the content where tones are the strongest
Personal and business communications:
Anyone could use the Tone Analyzer service to get
feedback about their communications, which could
improve the effectiveness of the messages and how
they are received.
Bloggers and journalists could use the Tone Analyzer
Service to get feedback on their tone and ﬁne-tune their
writing to reflect a speciﬁc personality or style.
Financial advisors and investors could use the Tone
Analyzer service to look at the tones reflected in
announcements and reports from the companies that
they are researching and investing in.
Automated contact center agent:
If a customer interacting with a support center is
agitated or angry, Tone Analyzer Service could be use
to detect those tones and quickly escalate the request.