This content is part of the Conference Coverage: A conference guide to AWS re:Invent 2021

Amazon Connect adds quick bot setup, call summarization

AWS' Amazon Connect CCaaS offering comes into sharper focus with new features for contact center managers, agents and data scientists setting up bots.

AWS' contact center as a service, Amazon Connect, added new features to bolster contact center agent productivity, more easily set up self-service bots and give contact center managers more automated insights into customer call transcripts.

AWS released a browser-based unified agent application, which aggregates services such as customer authentication, calls and chats with customer insights and knowledge articles into one interface. In preview is Amazon Lex Automated Chatbot Designer, which the company said reduces chatbot design from weeks to hours. It comes just two weeks after Google released its own quick-setup chatbot, "Bot-in-a-Box."

Chatbots still need fine-tuning and training once they're up and running, acknowledged Pasquale DeMaio, AWS general manager at Amazon Connect. What the quick-setup features do is automate the ingestion of data, as well as deploy AI to connect data about customer problems to the content that contains answers. For some users, it can eliminate weeks or months of work with third-party companies that set up chatbots, as well as the expense.

It can also potentially create more well-rounded bots, as Lex can ingest all of a user's data, while chatbots programmed by humans can only review and incorporate a finite amount of call transcript data, DeMaio said.

Pasquale DeMaio, AWS general manager at Amazon ConnectPasquale DeMaio

AWS also released automated call summaries via its Contact Lens machine learning transcription service. The feature identifies and tags key parts of a customer conversation -- examples of which might be "customer issue," "outcome" or "action item" -- and displays a summary that can be expanded to the full call transcript. Agents can revisit the summary when following up with a customer, and managers can retroactively understand the context of an interaction.

In aggregate, the features are "crowd-pleasers," said Dan Miller, founder of Opus Research.

"Some of them are directed at recently emerging competitors that have carved out things like call summary as a valuable feature of a contact center platform," Miller said.

Automated call summaries replace short surveys that call agents are often asked to fill out to help managers understand customer issues and how to solve them. Mining data in summaries is not only far more detailed than quick surveys, Miller said, but it enables agents to skip the questionnaires and move on to the next customer.

"It's addressing agent productivity and customer experience at the same time," Miller said.

Salesforce partnership, FedRAMP compliance expand

Amazon Connect started out life as a commoditized version of Amazon's homegrown contact center technology. As it grows in the market, its new features reflect mostly user requests, AWS' DeMaio said. While automation and self-service are foundational to the technology, the endgame is to improve the experience of human agents, not eliminate their jobs.

"When we think about automation, people think, 'Oh, you're just trying to replace the agent,' but that's not the case," DeMaio said. "In fact, the goal here is to remove annoying things the agents aren't great at and focus on where humans really bring benefit. You want to get their focus off of repetitive tasks. They're incredibly valuable for building that relationship with your customer."

In the last month, Amazon Connect potentially moved further out of the silo of AWS customers and into the mainstream contact center technology marketplace by adding FedRAMP Moderate certification and building deeper connections into the Salesforce platform. FedRAMP Moderate opens Amazon Connect to numerous federal, state and local government entities that have adopted the data security compliance standard.

AWS integrated Amazon Connect Voice ID to Salesforce, a voice biometric that validates a caller's identity. The Salesforce-AWS partnership, launched in 2016, added a big contact center connection when Salesforce chose AWS to host telephony for Salesforce Service Cloud Voice in 2019. The two companies have integrated many features since then that mingle the companies' data and automation capabilities, such as CRM routing that uses AI in Amazon Connect to detect customer intent and sends them to the right contacts in Salesforce Sales or Service Cloud voice, chat, email and SMS channels.

Salesforce, DeMaio said, uses Amazon Connect in its own contact center.

"Salesforce is a tremendous partner for us, and they're also a customer of Amazon Connect," DeMaio said. "Our teams work super closely together, and we're doubling down on that relationship because it's all based on what the [users] want."

Don Fluckinger covers enterprise content management, CRM, marketing automation, e-commerce, customer service and enabling technologies for TechTarget.

Dig Deeper on AI for customer service

SearchContentManagement
SearchUnifiedCommunications
SearchDataManagement
SearchEnterpriseAI
SearchERP
Close