The Growing Need for Knowledge Management

With artificial intelligence (AI) being hailed as the technology of the year, the role of knowledge management (KM) for contact centers has never been more important. Self-service, chatbots, virtual agents, and digital tools all rely on knowledge bases and knowledge management behind the scenes. In the new digital-first customer service environment, and with the rapid rise of chatbots, effective knowledge management tools are essential. 

A knowledge management platform is more than just a knowledge base – it’s used to capture, store, organize, and retrieve knowledge or information within an organization. Content includes articles, manuals, videos, FAQs, and other documents, and come from a variety of internal and external sources.

According to Verint, “At its core, knowledge management’s primary goal is to make organizations more efficient by championing access to existing documentation, experience, and ideas – whilst also capturing new knowledge. It does this by creating systems or processes that allow a workforce to share this information as a single source of truth, making knowledge available at the right time and in the right place.”

In contact centers, a knowledge management platform provides agents with quick and easy access to relevant information they need to assist customers, helping to ensure that agents provide consistent and accurate answers to customers.

Some of the key features of a knowledge management platform for contact centers include:

• Search capabilities to quickly find information

• Content creation tools for creating and updating information and knowledge

• Analytics tools to help managers track performance

Helping Agents Help Customers

Initially knowledge management platforms were used to better support agents, providing them with the information needed to better serve customers. In most enterprises, knowledge is spread out and scattered in multiple systems and hard to access. A knowledge management platform pulls together this information into a single repository, making it easy to search and surface the information needed quickly and easily. 

Rather than spending valuable time researching and looking up information that may be located in different databases, agents can access the knowledge and information on their desktop screen to effectively answer customer questions and resolve issues. This not only helps agents do their jobs better, it reduces agent training time, improves customer satisfaction, increases first contact resolution, and reduces interaction handle time, helping to lower costs and improve the bottom line.

In most cases, KM is used behind the scenes for agent assist, self service, FAQs, etc., and the KM is part of those solutions to support these capabilities. 

Helping Customers Help Themselves

More recently, thanks in large part to customer preference for self-service, these same knowledge management platforms can be used to support customers directly. Whether used in conjunction with chatbots, intelligent virtual agents, or simple FAQs, KM systems have become a critical part of the modern contact center for customer self-service. 

While the primary use case for KM is currently assisting agents, expect to see its usage for assisting customers directly increase in the coming year. 

I began hearing more and more from enterprise businesses about their growing use of knowledge management in contact centers in the past few years. In fact, when speaking with a group of customers at a Verint conference a few months ago, the role of KM came up repeatedly. When asked which Verint product they use the most or find the most helpful, they all mentioned knowledge management. I became intrigued and wanted to learn a bit more. To that end, I spoke with several contact center/CCaaS vendors about how they view and use KM. 

KM Options 

There are several vendors that have been offering knowledge management solutions for years, including Bloomfire, eGain, NICE, Puzzel, Verint, and others. These and other vendors provide KM solutions that work in conjunction with contact center/CCaaS offerings to provide complete solutions, enabling agents and customers to access the information they need.

Most of the contact center/CCaaS vendors have long-standing partnerships with KM vendors to integrate their solutions. For example, Avaya primarily partners with Verint, although it also has connectors to eGain and others; Cisco partners with eGain, Verint, and Google; Five9 provides connectors to eGain, Shelf, Google, IBM, etc. Genesys integrates with LivePro, Shelf, eGain, KMS Lighthouse, and others.

Over the past few years, there’s been a trend toward single suite solutions offering a fuller range of functionality. For example, more and more CCaaS vendors have added workforce engagement management (WEM) and workforce optimization (WFO) functionality to their contact center offerings. Many organizations, especially small- and medium-sized businesses, prefer reducing the number of vendors used and getting as much functionality from a single vendor as possible. This makes purchasing, deployment, and management simpler, while generally reducing costs. 

To that end, CCaaS providers are increasingly adding their own KM capabilities to their suite of offerings. For example:

• NICE CXone Expert knowledge management platform is a key part of the CXone platform and NICE’s digital self-service offerings. 

• Talkdesk offers Talkdesk Knowledge Management as part of its single suite solution. Talkdesk Knowledge Management manages information scattered across different locations by connecting multiple third-party platforms such as Salesforce, Confluence, Zendesk, and others. It uses AI-powered semantic search techniques to deliver contextual knowledge to agents through Talkdesk Agent Assist.

• Genesys added its own KM solution to its portfolio. Genesys Knowledge provides knowledge capabilities for digital, voice, multichannel, and employees/agents.

There’s no shortage of options for knowledge management offerings.

• eGain’s Knowledge Hub is part of its Customer Engagement Platform and is used to support a variety of eGain’s offerings including Virtual Assistant, self-service, Agent Assist, and the new Instant Answers, which leverages generative-AI technology. Knowledge Hub content can be personalized based on the specific channel and audience. eGain works with a large number of partners and shares presence on each others’ marketplaces. The company also provides out-of-the-box pre-built connectors for a large number of CCaaS, CRM, and other vendors, as well as APIs that can be used to create integrations based on eGain’s composable architecture.

• Verint's Knowledge Management uses patented, AI-infused search technology to create a more automated, natural way to connect people to knowledge. AI-infused Contextual Knowledge can be used to create a more automated, natural and effective way to connect people to answers. The information can be surfaced automatically and in real time across all channels to help agents respond more quickly without spending time searching for the right information. Verint has out-of-the-box adaptors and APIs, and partners with many CCaaS/contact center vendors, including Avaya, Five9, and others.

• Salesforce: Knowledge has been the core of service at Salesforce, as it is integral to case management and an important part of Salesforce’s offerings, including Service Cloud, Marketing Cloud, and Sales Cloud, and across all interaction channels. Salesforce notes that knowledge management should be integrated into the entire customer experience, providing a central source of knowledge and information for organizations to access when needed. Knowledge is central to how Salesforce customers shift to scale their service operations to be more efficient while improving customer experience. Knowledge allows agents to be more productive in resolving customer issues and end users to resolve issues effectively on their own terms. Salesforce focuses on the tight integration between knowledge and its applications, including chat, self service, etc. so that knowledge can be created once and deployed anywhere (in the console and across channels and workflows). By having out-of-the-box integration, customers don’t have to maintain multiple and separate solutions, reducing the effort to make everything work together. The company notes that knowledge is very connected to Salesforce’s story around data and bringing various sources of data together in a trusted and harmonized way.

By leveraging Salesforce Data Cloud, which ingests, harmonizes, and unifies customer data, information can be tailored based on what is known about the customer, providing a more personalized and customized level of response that is easily digestible. The Salesforce Data Cloud makes it easier to bring in various sources of information, including from Salesforce itself as well as external sources, and harmonizing the information to create a personalized experience. The external sources of data can be brought into the Salesforce platform and become part of the core data source that can be managed and trusted just like the internal information. The key is to ensure that the response based on the knowledge is personalized for the user, whether that’s an agent or end-user customer. For example, if an agent or customer accessing the knowledge is technical, they may want a more technical response. Similarly, the response can be customized based on the user’s experience level – if they’re new to the company and just ramping up or more experienced.

Contact Center Vendor Offerings

Here’s a quick update on how a few of the contact center/CCaaS providers I spoke with view KM, as well as their various approaches. 

• Avaya sees knowledge as an established part of the contact center used extensively by customers. Avaya’s job is to make sure the knowledge appears when the customer and agent need it. Working with its partner Verint, which provides the KM system, Avaya focuses on how the knowledge is presented to the agent or supervisor, noting that it’s all about composing the experience when the customer calls in to bring up relevant knowledge in composable experiences.

Avaya is focusing its strategy and product development and roadmap on the Avaya Experience Platform (AXP), a rebrand of its OneCloud CCaaS cloud-based contact center solution, with additional cloud capabilities. A key part of the Experience Platform is the unified desktop, integrated workspace, and widget framework, providing a converged unified client and contact center platform. AXP allows widgets to connect to the widget portal. Knowledge is presented as part of a widget and connect through APIs to any third-party knowledge solution. 

While Avaya and its customers mostly use Verint, AXP has open interfaces that lets Avaya connect to and embed any knowledge management that supports REST APIs. For knowledge management, contact center agents and supervisors currently use multiple interfaces, but will be able to use a single interface with the Experience Platform, where the Verint knowledge management will be added to the unified desktop as a widget. For AXP, Avaya has plans to use Verint KMPro, and Google CCAI. The next step for the Experience Platform is to include harvesting the data collected by the platform and turning it into actionable information for partners and end customers that go beyond simple performance metrics.

• Cisco notes that insight and knowledge are becoming more important, and that while the channels (voice, chat, email, etc.) are becoming commoditized, the information in these channels, and what you can glean from this information, have taken on new importance. Cisco has historically leveraged third parties like eGain to provide a formal KM solution for customers, and continues to go to market with both eGain and Verint for customers looking for a more complete solution. These solutions have been tightly integrated into Cisco’s Finesse desktop through existing APIs to provide KM solutions for existing contact center agents. 

In addition to these partnerships, Cisco has been enabling its own capabilities. Cisco’s acquisition of IMI mobile (now fully integrated in CCaaS as Webex Connect) came with significant capabilities in new messaging channels and automation. Each of these channels is capable of capturing insights that can be acquired, accumulated, and stored in a KM solution. Cisco captures this information through its Journey Data Service and exposes the data and events though APIs. In addition, Cisco plans to deliver topic analysis and machine learning models for post-call analytics in support of Webex Contact Center. This topic analysis modeling enables contact center business analysts to assess the reasons or call drivers for customer calls, and improvements in agent workflows. These in-house models are also planned to be exposed for third party KM/analytics engines. 

Cisco also has a go-to-market relationship with Google for Cisco Answers, using AI to pre-populate the agent desktop with insights from KM based on word and topic spotting observed in the engagement process. 

• Five9 notes that a well-designed knowledge management system is becoming increasingly important as customer interactions become more complex and the need for personalized and accurate information increases. KM systems enable virtual agents to access relevant information and respond accurately to customer queries while helping live agents provide timely and effective support.

Five9's strategy is to allow its customers to use their existing knowledge management systems. The company provides out-of-box integrations with the KM modules of the leading CRM vendors (Salesforce, Microsoft, ServiceNow, Oracle, Kustomer, Freshworks, Zoho, eGain, and Zendesk) and with Knowledge Management application vendors (Verint, eGain, Shelf.io) to provide customers with maximum choice. Five9 solutions, including IVA, Agent Desktop application, and Agent Assist, pull data from these customer KBs to provide relevant information to customers and agents across touchpoints.

AI tools like Agent Assist and IVAs use APIs to access and query various sources of knowledge bases and display them in real time. The key issue to address today is how to make it easy for customers to access and connect the various sources of knowledge. 

Five9 notes that it's all about the connections and is working on making it extremely easy to access information, minimize the number of clicks and present the information and knowledge to the agent, virtual agent, or customer.

• Genesys sees knowledge management as being foundational to delivering experience orchestration, an area of focus for the company. Knowledge management is needed to support Genesys’ experience orchestration, which is the coordination of technology, interactions, and touchpoints to enable organizations to capture data, personalize experiences, empathize with customers, predict, automate, engage, and learn. 

Genesys developed its own knowledge system called Genesys Knowledge, which has been available since the second half of 2021. Genesys acquired Bold360 at the end of 2021, providing the company with strong knowledge capabilities, mostly for digital interactions. In 2022 Genesys created knowledge capabilities for digital, voice, multichannel, employees/agents, and introduced the next version of Genesys Knowledge in November 2022. Genesys Knowledge is channel agnostic and surfaces contextual, intent-driven knowledge to customers and agents with AI-based semantic search. Customers can create their own articles, port articles from other databases, add content and context, etc., and then train the knowledge to be used with AI and automatically surface within the various channels. 

Genesys Knowledge is part of Genesys Cloud CX and is fully integrated into the Genesys platform. It is used in Agent Assist to bring up articles automatically. Agent Assist will also include automation for agents, including automated summarization, after-call summary, etc. The same knowledge base is used for bots, agents across channels and the Genesys knowledge portal, which enables self-service smart search. 

• NICE: CXone Expert, based on NICE’s Mindtouch acquisition, is used to meet customers at their point of need and to make the right self-service answers easy to find with smart knowledge management. CXone Expert is a key part of the new NICE FluenCX, a complete digital CX used to help organizations “Meet customers at the start of their journey: from web to mobile, from search to sites, and guide them to resolution.” FluenCX drives consumers to an organization’s knowledge content at the start of their journey, including search engines, making it easier for customers to find answers and information on their own via self-service digital channels. Expert is used to enhance self-service experiences with product and service knowledge and content to search engines, social and digital channels, chatbots, websites, and community forums. 

NICE recently announced the integration of CXone Expert with OpenAI's generative modeling used in ChatGPT to help organizations create human-like conversational consumer experiences without engaging agents. It leverages NICE Enlighten AI models and organization-specific data to create unique conversational AI models to provide self-service responses that are accurate and optimized for consumer understanding. CXone Expert can learn from past interactions, making it more accurate and effective over time.

NICE claims that it is “Ushering in a new era of CX, where consumers are immediately routed to the right answers with no need for transfers or call-backs, while creating exceptional self-service experiences that feel familiarly human.”

KM Challenges

Of course, there are challenges with many of today’s KM solutions, including:

• Supporting multiple touchpoints: Today’s systems are suitable for when agents interact with one customer at a time but may not be able to support multiple touchpoints. Interacting with multiple customers on web chat or SMS requires providing the same knowledge to agents on various channels. 

• Overwhelming agents: It’s important to provide the key articles without distracting the agent, and to surface the knowledge article at the right time. Providing summarized information of the knowledge and articles rather than the entire document has been a challenge, but advances in AI, notably GPT, will soon overcome this issue.

• Channels: It may be hard to provide the right level of knowledge based on the interaction channel. For example, if an agent is on a call with a customer, the knowledge and information provided needs to be succinct and easy for the agent to reference. For a chat interaction, the system should be able to show more information and articles. Depending on the channel, the knowledge management systems should be able to cater the response accordingly. 

• Accessing information that’s not in the knowledge management system, as “anything can be a knowledge article.” While KM makes it easy to access the knowledge articles, there are other knowledge sources, such as Salesforce and Sharepoint that may not be easily accessible. 

Looking Ahead

While KM systems are useful in and of themselves, going forward, I expect that KM systems will increasingly be purchased as part of larger solutions such as Agent Assist, virtual agents, etc. Customers are becoming more interested in the capabilities that KM enables. 

Of course, generative AI, such as ChatGPT, will have a huge impact on knowledge management. Generative AI can analyze large volumes of content, automatically categorize it based on intent, and convert data into consumable content. Expect to see generative AI used to generate knowledge articles and even rewrite knowledge content. One main issue is that ChatGPT was trained on general, publicly available information, and vendors will still need to add domain-specific information to supplement what ChatGPT provides.

eGain announced the general availability of eGain Instant Answers, making it easy for users to find relevant answer “snippets” from enterprise knowledge bases using generative-AI. Using LLMs that have been trained for enterprise-specific content, eGain Instant Answers finds the best answer snippets from longer documents. 
According to NICE, its OpenAI integration will enable the CXone Expert chatbots to create and deliver responses to customer self-service inquiries in a more human-like and conversational tone. The new Enlighten Actions brings together Enlighten, NICE’s AI for CX, with the OpenAI generative models to let organizations build AI-powered CX processes. Enlighten Actions make knowledge more accessible, enabling organizations to generate brand-specific actionable outputs. Enlighten Actions is integrated across NICE’s portfolio of products, including the CXone Expert knowledge management solution, CXone’s Bot Builder, and SmartAssist.

Salesforce has been early to the game, layering generative AI onto Einstein for Service and Customer 360 to automatically generate personalized responses for agents to quickly email or message customers. Einstein GPT for Service will be used to generate knowledge articles from past case notes and personalized agent chat replies.
Salesforce notes, “We’ll be able to train the AI across all of the case related information and conversations captured in Salesforce to automatically generate drafts of knowledge articles for human review, drastically cutting the time to create knowledge and making it easier to keep articles up to date.” The current agent-assist tools within Service Cloud listen to customer conversations, predict which knowledge article or piece of CRM data may prove helpful, and surface it for agents to help provide a response to the customer inquiry or issue. When used with Einstein GPT, the Service Cloud will be able to auto-generate the entire response for agents, which they can review or approve, and publish. Einstein GPT can also auto-generate knowledge articles for evaluation based on successful customer conversation transcripts. The key for the near future is to maintain the human-in-the-loop to safeguard trust, and ensure accuracy and appropriateness.

While Generative AI will make knowledge more accessible for self-service capabilities, organizations need to ensure the accuracy of the information. ChatGPT can do amazing things, but it is also known to produce incorrect information and “hallucinations.” Human oversight will be required for the foreseeable future, reviewing the responses to ensure accuracy and non-bias.

We’re in the very early days of generative AI and ChatGPT so it’s too soon to predict its impact on knowledge management, but expect it to be significant.

Conclusion

Knowledge is a crucial element for both self- and assisted service in contact centers, requiring a modern, adaptable knowledge management platform. Organizations have a variety of options, with a variety of vendors offering these platforms. ChatGPT and generative AI will have a dramatic impact on knowledge management systems in the near future, impacting the way knowledge is curated, accessed, and used.