In this session, you will learn best practices for implementing simple to advanced use cases on Amazon Connect - a self-service, cloud-based contact center service that makes it easy for any business to deliver better customer service at lower cost. Amazon Connect is based on the same contact center technology used by Amazon customer service associates around the world to power millions of customer conversations
2. Session Objectives
• Introduction to AI Services
• Introduction to Amazon Connect
• Walkthrough a Use Case and design considerations for
building conversational interfaces
• Historical Reporting and Analytics
Amazon Lex AWS
Lambda
Amazon
Connect
Amazon
SNS
Amazon
S3
Amazon
Kinesis
Amazon
Athena
Amazon Polly
5. Amazon Rekognition
Deep learning-based image recognition service
Search, verify, and organize millions of images
Object and Scene
Detection
Facial
Analysis
Face
Comparison
Facial
Recognition
10. Amazon Lex: Conversational Chatbots
I’d like to book a hotel.
Sure! What city would you like to book it in?
11. Amazon Lex: Conversational Chatbots
I’d like to book a hotel.
Sure! What city would you like to book it in?
New York City
Destination:
New York City
12. Amazon Lex: Conversational Chatbots
I’d like to book a hotel.
Sure! What city would you like to book it in?
New York City
When would you like to check in?
Next Friday
Destination:
New York City
Check In:
21/7/2017
13. Amazon Lex - Technology
Amazon Lex
Automatic Speech
Recognition (ASR)
Natural Language
Understanding (NLU)
Same technology that powers Alexa
Cognito CloudTrail CloudWatch
AWS Services
Action
AWS Lambda
Authentication
& Visibility
Speech
API
Language
API
Fulfillment
End-Users
Developers
Console
SDK
Intents,
Slots,
Prompts,
Utterances
Input:
Speech
or Text
Multi-Platform Clients:
Mobile, IoT, Web,
Chat
API
Response:
Speech (via Polly TTS)
or Text
14. Lex: Build Natural, Conversational Interactions In Voice & Text
Voice & Text
“Chatbots”
Powers
Alexa
Voice interactions
on mobile, web
& devices
Text interaction
with Slack & Messenger
Enterprise
Connectors
(with more coming) Salesforce
Microsoft Dynamics
Marketo
Zendesk
Quickbooks
Hubspot
17. AI Services
AI Platform
AI Engines
Amazon
Rekognition
Amazon
Polly
Amazon
Lex
More to come
in 2017
Amazon
Machine Learning
Amazon Elastic
MapReduce
Spark &
SparkML
More to come
in 2017
Apache
MXNet
TensorFlow Caffe Theano KerasTorch CNTK
Amazon AI: Democratized Artificial Intelligence
18. AI Services
AI Platform
AI Engines
Amazon
Rekognition
Amazon
Polly
Amazon
Lex
More to come
in 2017
Amazon
Machine Learning
Amazon Elastic
MapReduce
Spark &
SparkML
More to come
in 2017
Apache
MXNet
Caffe Theano KerasTorch CNTK
Amazon AI: Democratized Artificial Intelligence
TensorFlow
P2 ECS Lambda GreenGrass FPGAEMR/Spark
More to
come
in 2017
Hardware
20. Tens of thousands of Customer Service Associates support Amazon
customers around the world.
Amazon strives to be
Earth’s most customer-centric company
AMAZON
SUPPORTS
Millions Of Customers
Dozens Of Languages
32 Countries
21. Solutions in the marketplace
couldn’t meet our expectations
Heavy
Professional
Services
Hardware,
Telephony
Integration,
Space
Complex
Pricing
Security,
Scalability,
Reliability
Difficult
Integrations
Cumbersome
Tools
We needed the right contact center technology to deliver on
our promise
23. Real time and
historical analytics
Skills-based routing
[Automatic Call Distribution (ACD)]
Call
recording
High-quality
voice capability
Easy to use, cloud-based contact center solution that scales
to support businesses of any size
With Tools That Grow With Your Needs
Amazon Connect
27. Ok, you are now
booked for a
9:00AM departure
tomorrow out of
San Francisco,
arriving in Seattle
at 11:45AM.
Can you
please
rebook me
for the
same flight
tomorrow?
Great
Thank you!Data
Dip
CRM
content
Hi Nikki Wolf,
I see your flight
was cancelled
today. How can
I help you?
Incoming
customer
call
Contact Flow Engine – Customer Experience
28. NATURAL
Amazon Lex Chatbots
use the same technology
that powers Alexa
DYNAMIC
Answer customer
questions before they
are even asked
PERSONAL
Contact flows adapt on
a per customer basis
Ok, you are now
booked for a
9:00AM departure
tomorrow out of
San Francisco,
arriving in Seattle
at 11:45AM.
Can you
please
rebook me
for the
same flight
tomorrow?
Great
Thank you!Data
Dip
CRM
content
Hi Nikki Wolf,
I see your flight
was cancelled
today. How can
I help you?
Incoming
customer
call
Contact Flow Engine – Customer Experience
30. OPEN PLATFORM
Your
S3 Storage
Your Data
Warehouse
Customer
Databases
Business
Intelligence
Workforce
ManagementAgent Data
AWS
Lambda
Call
Recordings
Metrics
Contact
Flows
CRM Contact
Control Panel
Open Platform/ Easy Integrations
33. Telephone numbers available in these countries:
Australia
Austria
Belgium
Canada
Czech Republic
Denmark
Finland
France
Germany
Greece
Hong Kong
Ireland
Italy
Japan
Mexico
Netherlands
New Zealand
Norway
Poland
Portugal
Slovakia
Spain
Sweden
Switzerland
United Kingdom
United States
• New with APAC region support (Sydney)
• New with EU region Support (Frankfurt)
Phone Number Availability
35. Example Use Case
• Key requirements
• Receive Call to a Team Member
• If the Agent is unavailable forward to Team
• Escalate calls across teams
• Call Recording with Search
• Outbound click to dial from Salesforce
• Email missed call to Team
Amazon
Connect
50. Summary of Configuration
Queues
• Agent1Queue
• Agent2Queue
• Team1Queue
On each queue, set an outgoing
number dedicated to individual
users.
Amazon
Connect
Working Profiles
• Agent1Profile
• Agent1Queue
• Team1Queue,
• Set the outgoing queue to
Agent1Queue.
• Agent2Profile
• Agent2Queue
• Team1Queue
• set the outgoing queue to
Agent2Queue.
Users
• Agent1
• Agent2
51. Salesforce Integration
• Install a Package in Salesforce
• Provide a configuration xml
• Connect Application Configuration
• Support Click To Dial
Amazon
Connect
52. Amazon Lex Integration
• Give Connect Permissions to
post content to Lex
• E.g. add an in line policy to
connect-access-role-xxx
Amazon
Connect
53. Create Amazon SNS Topic
• Create a topic and note arn: for AWS Lambda
• Subscribe to topic with an email group
Amazon
SNS
54. Build AWS Lambda Function
• Create a role
AWS
Lambda
Create a Lambda Function
55. Build AWS Lambda Function
• Declarations
• Function close –
response for
Amazon Lex
AWS
Lambda
56. Build Lambda
• Intent leave_message Read slots passed from Lex
builds the message and send to SNS topic
• (Store ARN as Lambda variable)
AWS
Lambda
57. Build Lambda
The handler calls the dispatch, checks the intent name
from Lex and then calls the function to process
AWS
Lambda
65. Lex – Intent LeaveMessage
Utterances:
The question is asked “Would you like to leave a message?”
Amazon Lex
66. Lex - Slots
MyName: The Name of the Person
CallReason: CallReasons Reasons for call
Date
Time
CompanyName: Name of the organisation
Amazon Lex
67. Lex - Fulfilment
Call the Configured Lambda. This will send the json format
containing the slot values from the intent.
Amazon Lex
68. Add Other Intents
Add the following intents using the associated Built In Intents
• Cancel using AMAZON.CancelIntent
• Stop using AMAZON.StopIntent
Create Intent “No”
Create Intent “ExitCall”
Amazon Lex
69. Lex - Testing
• Build and Test
• If there is an error check
CloudWatch log files for Lambda.
• Publish and choose an alias
Amazon Lex
70. Configure Connect to Integrate with Lex
In Contact Flows
+Add Lex Bot
Select the Lex Bot
MailBox
Amazon
Connect
71. Connect Lex Call
In Contact Flow
In the contact flow “Get
Customer Input”
Set the Text for Connect to
prompt
Set the name and alias for
Lex (they are case sensitive)
Add the intents
• LeaveMessage
• ExitCall
• No
• Stop
• Cancel
Amazon
Connect
72. Call from Lambda
In Contact Flow
On Invoke AWS Lambda
Add the ARN of the Lambda
Function
ConnectMessageSender
This flow will email the
contact details to the team
for calls received where a
message was not taken
Amazon
Connect
79. Example Contact Trace Record
Amazon
S3
• JSON records stored for
contact flows
• Full context of call
recordings can be
queried directly from S3
repository
80. Configure Athena to Read Data
Amazon
Athena
• Create a database “connectreports”
81. Configure Athena to Read Data
Amazon
Athena
• Creating tables in Athena https://aws.amazon.com/blogs/big-data/create-tables-in-amazon-athena-from-nested-json-and-
mappings-using-jsonserde/
• Supporting tool to assist conversion https://aws.amazon.com/blogs/big-data/create-tables-in-amazon-athena-from-nested-
json-and-mappings-using-jsonserde/
• Cataloguing tables with AWS Glue http://docs.aws.amazon.com/glue/latest/dg/add-crawler.html
86. Summary
• Contact Flows and Call Escalation
• Call Recordings Stored in S3
• Integration with Lex/Polly for Conversational interfaces
• Integrate with Kinesis for ingestion of Contact Trace Records to S3
• Analytics and reporting tools
• Click to dial from Salesforce
• Abandoned calls are processed by Lambda function
Focus on Voice Quality and Pronunciation
Automatic Accurate Text Processing
Intelligent and Easy to Understand
Add semantic meaning to text
Customized Production
A viseme represents the position of the face and mouth when saying a word. It is the visual equivalent of a phoneme, which is the basic acoustic unit from which a word is formed. Visemes are the basic visual building blocks of speech.
Support for lip sync animation -
So Lex has identified the context for the user, and can guide the user through steps so that they provide the information required to book this hotel.
Lex now knows where we’re going.
We can then pick out “Friday” and work out the date.
At Amazon, we strive to be Earth’s most customer-centric company.
And to do this, everyone who works at Amazon understands that our customers are our top priority and that our future as a company is built on our customers’ satisfaction.
Today, Amazon supports millions of customers who speak dozens of languages from 32 Customer Service Centers across the globe.
Consider that during the 2016 holiday season peak online shopping period, tens of thousands of Customer Service associates supported Amazon customers around the world.
At Amazon, we consistently strive to ‘raise the bar’ for our customers, and our search found that the traditional call center solutions were unable to meet our demanding requirements.
The tools were cumbersome and required highly technical and niche expertise.
The integrations were complex with multiple components, channels, and vendors.
The systems were built in proprietary environments requiring esoteric skills and heavy professional services.
We were also looking for increased security, scalability, and reliability.
Nuggets:
On Prime Day, for example, we need to be able to scale to meet the high volume of customer interactions in a single day, but be able to contract back down quickly.
At Amazon, reliability is second only to security. The solution we sought needed to be secure, but provide the highest level of availability in order for us to serve our customers and their needs.
Available solutions did not meet our needs for space, hardware, and telephony integration.
Finally, the pricing models were complex, requiring long-term commitments.
To deliver on our goal to be Earth’s Most Customer-Centric Company, we knew we needed a better way for our agents to engage with our customers.
So we built it!
Using our knowledge and experience, we built an easy-to-use, intuitive system that enables our customer service agents to focus their attention on the customer, rather than on navigating a complex, difficult-to-use system.
Over 50 groups and businesses within Amazon and its subsidiaries provide customer service today, including Zappos, Audible, and of course, Amazon Retail’s customer service.
Amazon Connect is the contact center solution we designed and built to enable Amazon to deliver an exceptional customer experience.
It includes all the standard functionality contact center operations teams have come to expect.
Skills-based routing maximizes the efficiency of agents and the satisfaction of customers.
Calls are recorded, making it easy to monitor and improve agent quality.
All the common reports and analytics you would expect are available and easy to use, including Service Level, Handle Time, and agent performance.
And high-quality voice comes standard with Amazon Connect.
Beyond the standard functionality, Amazon Connect delivers differentiators that allow AWS customers to create exceptional customer experiences.
For example, Amazon Connect can be setup in minutes and agents can take calls after just a few simple steps.
Optional Slide:
Amazon Connect comes ready with an intuitive, self-service setup. You’ll be taking your first call after a few simple steps!
And there are no complex deployment or hardware integrations required.
Amazon Connect is intuitive and easy-to-use, allowing agents to be trained quickly. And your agents can focus on your customer, rather than navigating a cumbersome system.
Nugget: Other solutions can take 6-18 months to setup and deploy.
The Contact Flow Engine is dynamic and personal. By integrating with your customer data, you can anticipate end-customer needs, predicting and delivering answers to questions before they are even asked.
And you can design conversational interactions that feel natural to your customers, by integrating with Amazon Lex, giving you access to the same speech recognition and natural language understanding technology that powers Alexa.
Here is an example of Amazon Connect’s Contact Flow Engine enabling dynamic, personal and natural customer interactions.
And integration with Amazon Lex means customers can experience the same Speech Recognition and NLU technology that powers Amazon’s Alexa.
Customers get a better experience, and the entire interaction can be serviced completely within the contact flow.
Nuggets: Using Customer Data, Contact Flows can be created to identify the customer and determine how best to treat the interaction and how and where to route the call.
Amazon Connect is an open platform that is simple to integrate with other enterprise applications and AWS services like Amazon S3 & Lambda
Amazon Connect is simple to integrate with other applications, CRMs or databases.
AWS Lambda gives you the flexibility to run code for virtually any type of application or backend service.
And your data is your data. For example call recordings are stored in your S3 bucket, where you can control access and the lifecycle of your data using the powerful S3 tool set. [Build]
Also, you can stream Amazon Connect’s most detailed contact records to the data warehouse of your choice, and analyze as you choose with your tools of choice, like Amazon QuickSight or Tableau.
…And you can leverage AWS services or take advantage of the full AWS ecosystem to further enhance and customize your solution.
Amazon Connect can leverage the full ecosystem of AWS ISV consultancy partners to customize as needed.
Amazon Connect delivers a complete contact center at a fraction of the cost of traditional contact center solutions.
Like all AWS services, Pricing for Amazon Connect is pay-as-you-go and there are no required long term commitments.
You can instantly scale up or down to meet your needs with out worrying about telephony, hardware, space, or capacity.
Most importantly, with customer-connected-minutes, you only pay for the time your customers are connected to the service plus any telephony you use. There are no esoteric formulas based on agents, managers, capacity, or features.
Today, Amazon Connect instances can be created in North America (East), Frankfurt, and Sydney Regions. Full pricing details and country phone number availability can be found on https://aws.amazon.com/connect/pricing/
1. The Amazon Connect “per customer connected minute” pricing is simple and consistent across all regions ($0.018)
2. Telco pricing varies based on region, countries served, and type of numbers required (DID or Toll Free) from those countries.
3. Today, Amazon Connect instances can be created in North America (East), Frankfurt, and Sydney Regions. Full pricing details and country phone number availability can be found on https://aws.amazon.com/connect/pricing/
Other info:
(DID = Direct Inward Dial)
Taxes, surcharges, and fees may apply
1. The Amazon Connect “per customer connected minute” pricing is simple and consistent across all regions ($0.018)
2. Telco pricing varies based on region, countries served, and type of numbers required (DID or Toll Free) from those countries.
3. Today, Amazon Connect instances can be created in North America (East), Frankfurt, and Sydney Regions. Full pricing details and country phone number availability can be found on https://aws.amazon.com/connect/pricing/
Other info:
(DID = Direct Inward Dial)
Taxes, surcharges, and fees may apply
1. The Amazon Connect “per customer connected minute” pricing is simple and consistent across all regions ($0.018)
2. Telco pricing varies based on region, countries served, and type of numbers required (DID or Toll Free) from those countries.
3. Today, Amazon Connect instances can be created in North America (East), Frankfurt, and Sydney Regions. Full pricing details and country phone number availability can be found on https://aws.amazon.com/connect/pricing/
Other info:
(DID = Direct Inward Dial)
Taxes, surcharges, and fees may apply
When not present this was out of hours and left a message by Lex which emailed the Agent.