AWS has released new generative AI-powered features for Amazon Connect, its CCaaS platform.
Each new feature builds on existing applications within Amazon Connect and comes at no additional cost until March 2024.
Moreover, they leverage the LLMs available within Amazon Bedrock – which hosts models from various big-name vendors, such as Anthropic, Cohere, and Meta.
AWS teased many of these in September when highlighting how generative AI (GenAI) could improve agent-assist, manager-assist, and customer self-service experiences.
Now, it has packaged these up and hard-launched four of them during the AWS re:Invent conference – as unpacked below.
Amazon Q In Connect
AWS has released a new, customizable GenAI-powered virtual assistant – “Amazon Q” – across its tech stack.
Q allows employees throughout the organization to gain fast, relevant answers to pressing questions, solve problems, auto-generate content, and take action.
Those employees include contact center agents, who can now leverage Amazon Q in Connect.
Indeed, agents can chat with Q and uncover the necessary answers to respond to customer queries quickly without searching through various knowledge bases and documents.
Moreover, Q tunes into voice and digital interactions, detects customer intent, auto-suggests responses to queries, and auto-recommends actions for agents to further support customers.
Q connects with various contact center knowledge repositories – which agents typically use when conversing with customers – to bring these use cases to life.
In addition, the virtual assistant will cite any piece of information it leverages in its answers, so agents can dig deeper if they’d like.
Generative Post-Contact Summarization
Amazon Connect Contact Lens is the conversational intelligence platform natively available on the CCaaS platform.
AWS has now augmented this with GenAI to surface new insights from every contact center conversation – including an automated summary of what happened during the call.
Contact Lens condenses that summary into five or six crisp sentences while pinpointing any pending follow-up items and critical moments during the conversation.
Those critical moments help to transform manual call evaluation processes, optimizing the time supervisors spend searching for agent performance improvement and praise opportunities.
Yet, the biggest potential benefit of this capability is in speeding up post-call processing by streamlining the CRM work agents perform after each interaction.
After all, they may simply review and submit the automated summary to shave seconds from every customer conversation.
Lastly, it’s critical to note that these summaries follow the same format, making it easier to run analytics and garner insight from contact center interactions.
Unified Connect Customer Profiles
Creating unified customer profiles – which offer a single view of customer conversations, preferences, and purchases – is a longstanding contact center challenge.
Indeed, organizing and pooling various data flows into a single location and a unified format often takes weeks – if leaders are lucky.
Thankfully, AWS had already released Amazon Connect Customer profiles to simplify their creation and pull mission-critical customer information into the heart of the CCaaS platform.
Now, it leverages LLMs to further lessen the time contact centers take to build these profiles.
So, when an administrator adds data sources from numerous platforms – including Adobe Analytics, Amazon S3, Salesforce, ServiceNow, and Zendesk – to the Connect Customer Profile solution, it will automatically analyze it.
In doing so, it determines how to organize and combine data from multiple sources and funnel them into comprehensive, accurate profiles.
With this, managers can review, make edits, and complete the setup of customer profiles “with just a few steps” – according to AWS.
Lex Interprets Customer Utterances with Greater Accuracy
Amazon Lex is the conversational AI solution that sits inside Amazon Connect and is also available as a standalone platform that plugs into third-party environments.
Like many similar platforms, it has struggled to comprehend numeric values, creating escalations for many seemingly routine customer queries.
For instance, consider businesses that build bots on the platform to schedule appointments. A customer may ask: “I want to change my reservation for myself and my two children.”
Many conventional bots struggle to interpret how many people are on the reservation, which is – in this case, three – despite the customer writing two.
Now, with a new LLM-enabled “assisted slot resolution feature” inside of Lex, bots will understand such user prompts with much greater accuracy – improving containment.
Yet, the solution isn’t strictly for numbers. It also improves a virtual agent’s understanding of dates, phone numbers, confirmations (yes/no), and geographies.
As an example of the latter, if a customer said “Big Apple”, the response would run through an LLM, and the bot would correctly interpret it as New York.
While this is a powerful use case for GenAI-powered virtual agents, it’s not the first AWS has released for Lex – having previously made moves to increase the scope of contact automation with the tech.
Elsewhere at AWS re:Invent
In the past year, AWS has continued its march to the forefront of the CCaaS market after achieving “leader” status in the latest Garner Magic Quadrant.
In accomplishing this feat, AWS maintained a rapid pace of customer-led innovation, which it has continued in 2023, far beyond the generative AI.
That includes co-innovation with Salesforce, which will now pull Amazon Connect features into Service Cloud, and its new Thin Client – both announced earlier this week.
Yet, additional announcements keep flooding in with the launch of in-app, web, and video capabilities within Amazon Connect and two-way SMS.
The former helps businesses embed video customer service across web and mobile applications – which the agent or customer can enable (or perhaps both!).
Meanwhile, two-way SMS allows for back-and-forth customer communication across the channel without the need for costly third-party integrations that often silo data.
From a Visionary to CCaaS Leader: More Recent Enhancements to Amazon Connect
These innovations add a glazed cherry on top of the Amazon Connect cake, a recipe that AWS will continue improving every week – far beyond the re:Invent showstopper.
Indeed, in the summer, AWS added new analytics, chat, and routing capabilities to the CCaaS platform – with the former vastly enhancing the Connect’s reporting features.
After, the vendor released another stream of innovations, with standouts including “agentless” voice dialing and an agent desktop upgrade that allows Connect users to plug-in third-party apps.
As a result of the latter, contact centers can lower the number of applications agents switch between when handling customer interactions, enhancing agent efficiency.
In releasing such a reliable flow of upgrades to Connect, AWS achieved leader status in several industry studies – alongside the Magic Quadrant.
Check out our rundowns of some of these CCaaS reports below:
- The Forrester Wave for CCaaS 2023
- Gartner Peer Insights “Voice of the Customer” for CCaaS 2023
- The G2 Enterprise Grid Report for Contact Center, Summer 2023
