AWS Contact Center
Salesforce Contact Center with Amazon Connect: Streamlining omnichannel customer engagement
Introduction
Salesforce Contact Center with Amazon Connect (SCC-AC), now generally available, is a groundbreaking offering that integrates Amazon Connect native digital and voice capabilities into Salesforce. Building upon the existing voice-only Service Cloud Voice (SCV) integration, SCC-AC enables customers to unify voice and digital channels across Amazon Connect and Salesforce enhancing customer and agent experience as well as operational efficiency.
Key features
Salesforce Contact Center with Amazon Connect (SCC-AC) makes Amazon Connect channels and capabilities natively available in Salesforce. SCC-AC as a managed package is available in Salesforce App Exchange. In order to install the package in the Salesforce organization, the underlying Salesforce instance should be Salesforce Service Cloud Voice enabled. In the first phase, the following key features will be unlocked via SCC-AC.
New channels
SCC-AC offers native Amazon Connect Chat (SMS, web chat, social messaging). Customers can use Amazon Connect Chat with other features, such as Amazon Q in Connect (QiC) for self-service and Amazon Connect Contact Lens for conversational analytics.
Note: Amazon Connect Chat is currently available only with Service Cloud Voice Partner Telephony from Amazon Connect also known as Service Cloud Voice Bring Your Own Amazon (BYOA).
The following diagram explains high-level architecture, data flow, and API sequence for Amazon Connect messaging channels in context of SCC-AC.
- Customer initiates chat message from chat client hosted on website or mobile app or via SMS.
- Associated Amazon Connect contact flow is triggered to offer initial greeting and then self-service using Amazon Lex.
- Before transferring the chat to an agent, Create Messaging Session AWS Lambda function is invoked from the contact flow.
- Create Messaging Session AWS Lambda function invokes Salesforce Interaction Service API to create messaging session in Salesforce backend.
- After identifying agent, SCC-AC Connector receives routing request from Amazon Connect routing.
- SCC-AC connector fetches the messaging session from Salesforce backend using work item ID.
- Based on the messaging session, SCC-AC connector creates agent work in Salesforce backend.
- Salesforce backend routes the agent work to agent interface (Salesforce Omni-Channel) with customer message and screen popup.
- Once the agent types and clicks to send the message, SCC-AC connector receives the message via Salesforce API.
- SCC-AC connector sends the message to Amazon Connect using Amazon Connect Streams API.
- Amazon Connect sends the message to chat client.
Expanding voice channel
SCC-AC expands the voice channel from telephony to web and in-app voice calling, allowing end-customers to initiate a contact without leaving the website or mobile application. Organizations can pass contextual information to Amazon Connect to better personalize the customer experience based on customer information stored in Salesforce or based on actions previously taken within the app.
Unified contact routing
SCC-AC provides a unified routing engine across Amazon Connect and Salesforce channels. Such architecture helps simplify routing management, unify self-service, and improve reporting. For example, customers can route the Salesforce cases through Amazon Connect routing engine.
The following diagram explains the high-level architecture, data flow, and API sequence for unified contact routing in SCC-AC.
- The case is created in Salesforce, which could be using ‘email to case’ feature of Salesforce.
- This case triggers Salesforce flow (provided with SCC-AC package).
- Salesforce flow triggers Amazon Connect contact flow to create Amazon Connect Task with case id (other case details) as input attributes.
- Amazon Connect contact flow, based on the case information, creates and routes the Task, based on routing rules, set in Amazon Connect.
- SCC-AC connector receives routing request and creates agent work in Salesforce backend.
- Salesforce backend routes agent work with case details to the Salesforce Omni-Channel agent interface with screen pop.
Prerequisites
Salesforce Contact Center with Amazon Connect is a managed package available in the Salesforce AppExchange. In order to install the package in the Salesforce organization, the underlying Salesforce instance should be Salesforce Service Cloud Voice enabled (licensed) with Salesforce Digital Engagement (licensed).
Architectural choice
One of the key principles of SCC-AC is to provide customers a choice to select messaging channels and related features from either Amazon Connect or Salesforce. Voice channels – telephony and web and mobile app, are always native to Amazon Connect. This gives us two key reference architectures – Amazon Connect-oriented architecture and Salesforce-oriented architecture.
Amazon Connect-oriented architecture
This is an architecture choice if a customer decides to use native Amazon Connect messaging channels and other linked Amazon Connect features for self-service, agent assist and manager services. The agent and supervisor experience and customer data will always be native to Salesforce. The voice channels – telephony, web and mobile app will always be native to Amazon Connect.
Salesforce-oriented architecture
This is an architecture choice if a customer decides to use native Salesforce messaging channels and other linked Salesforce features for self-service, agent assist and manager services. The agent and supervisor experience and customer data will always be native to Salesforce. The voice channels – telephony, web, and mobile app will always be native to Amazon Connect.
Key benefits
Based on the architectural choices, customers unify and simplify contact center design, architecture, and feature map.
The following diagram shows some of those benefits with SCC-AC GA launch.
1.Unified self-service layer – In the Amazon Connect-oriented architecture (i.e. all the channels are Amazon Connect native), Amazon Q in Connect will be a single generative AI-powered self-service layer across channels. This will help create a unified self-service customer experience.
2.Unified conversation analytics layer – In the Amazon Connect-oriented architecture, Contact Lens will be a single conversational analytics layer across channels which enables various conversational analytics features:
-
- Transcription: Transcribing conversation between customer and agent in real time, which can be leveraged in different use cases.
- Categorization: Contact center administrators can set up Contact Lens rules based on spoken words/phrase, sentiment, and contact attributes. Contact Lens applies these rules in real-time conversations trigger actions based on those rules, such as Next Best Action in Salesforce for agent assist.
3.Unified routing engine – SCC-AC provides mechanisms to route Salesforce objects such as cases through Amazon Connect routing. This helps to unify and simplify routing configurations and improve reporting.
4.Native Salesforce experience – Agent and supervisor experience across Salesforce and Amazon Connect channels will be Salesforce-native. For agents and supervisors, this reduces training and learning time and also enables agents and supervisors to function from Salesforce console instead of switching between consoles.
Conclusion
Salesforce Contact Center with Amazon Connect (SCC-AC) seamlessly integrates Amazon Connect native channels into Salesforce. This enables organizations to create unified customer and agent experience with streamlined contact center operations.
Call to action
Authors
 |
Chintan Gandhi – Specialist Solutions Architect for Amazon Connect and Salesforce Integrations based in London, United Kingdom. Apart from helping our largest customers and strategic partners with new technology adoption and scaling, Chintan enjoys spending time with his family, playing chess, cricket and reading. |
 |
Steve Earl – Specialist Solutions Architect for Amazon Connect and Salesforce Integrations based in North Carolina. When he’s not helping customers, partners, and internal teams with customer experience, Steve enjoys spending time with his family, both human and canine, working on his mountain cabin, reading, and golf. |
 |
Marc Rudkowski – Principal Product Manager for Amazon Connect Partner Solutions based in Raleigh, North Carolina. He has worked with our largest Enterprise customers and partners to enable their CX journey in Amazon Connect since its launch in 2017. Outside of work, Marc is an avid cyclist and enjoys spending quality time with his family |