Why Your Customers Want a Cognitive Call Center
Download as PPTX, PDF1 like1,616 views
Customers increasingly expect to engage with brands through self-service channels and not with tedious and frustrating traditional call centers that use IVR. They want the convenience and timely, streamlined interactions that a virtual agent offers. More than 50% of enterprises have invested in virtual agents for customer service, and it’s estimated that by 2020 approximately 85% will manage the customer relationship with no human interaction at all. But not all virtual agents are created equal. IBM Watson harnesses the power of natural language processing, machine learning, and cognitive computing to deliver an exceptional virtual agent experience. Perficient and IBM took a closer look at intelligent virtual agents, including: -The benefits of intelligent virtual agents -Considerations when selecting a virtual agent -IBM Watson Assistant introduction and demonstration -Practical ways to get started with Watson
Ad
More Related Content
What's hot (20)
Similar to Why Your Customers Want a Cognitive Call Center (20)
Ad
More from Perficient, Inc. (20)
Ad
Recently uploaded (20)
Why Your Customers Want a Cognitive Call Center
- 2. 2 About Perficient Perficient is the leading digital transformation consulting firm serving Global 2000 and enterprise customers throughout North America. With unparalleled information technology, management consulting, and creative capabilities, Perficient and its Perficient Digital agency deliver vision, execution, and value with outstanding digital experience, business optimization, and industry solutions.
- 3. 3 Perficient Profile • Founded in 1997 • Public, NASDAQ: PRFT • 2017 revenue $485 million • Major market locations: Allentown, Atlanta, Ann Arbor, Boston, Charlotte, Chicago, Cincinnati, Columbus, Dallas, Denver, Detroit, Fairfax, Houston, Indianapolis, Lafayette, Milwaukee, Minneapolis, New York City, Northern California, Oxford (UK), Phoenix, Seattle, Southern California, St. Louis, Toronto, Washington, D.C. • Global delivery centers in China, India and Mexico • 3,000+ colleagues • Dedicated solution practices • ~95% repeat business rate • Alliance partnerships with major technology vendors • Multiple vendor/industry technology and growth awards
- 4. 4 • Practice Overview • Virtual Agents Trends and Forecast • Cognitive Virtual Agent Review ⎼ Conversational Agents ⎼ Conceptual Architecture • Training Approach • PoT Execution ⎼ Sprint Plan ⎼ Assumptions • Perficient Case Studies Agenda
- 5. 5 • Over 30 delivery professionals • Leveraging experience in Analytics, Big Data, Unstructured Content Management, Enterprise Search, Digital Experience and Business Optimization • IBM Watson Talent Partner • IBM Watson • Microsoft Azure • Google Cloud Platform • Amazon • Open Source (R, Python, etc.) • Platform Selection Engagements • Cognitive Readiness Evaluations • Solution Business Case Development • Cognitive Search Implementations • Text and Content Analytics Solutions • Virtual Agents and Chatbots • Predictive Modeling • Machine Learning Models • Decision Support Solutions Practice Overview Platform Support Offerings and Services 2017 Beacon Award Winner for an Outstanding Watson Cognitive Solution Artificial Intelligence Practice Overview
- 7. 7 Looking Forward by 2022 25% of customer service and support operations will integrate virtual agent technology across engagement 40% of customer-facing employees and government workers will consult daily an AI virtual support agent for decision support over 50% of organizations have already invested in VAs for customer service, as they realize the advantages of automated self-service 20% of brands will abandon their mobile apps in favor of building presence in consumer messaging apps, such as Facebook Messenger 85% of the enterprise relationship to a customer will be managed without human interaction 30% of all B2B companies will employ artificial intelligence (AI) to augment at least one of their primary sales processes. by 2019 in 2018 by 2020
- 8. 8 Self-Service Channels are Key to Winning the Future of Customer Service
- 9. 9 The volume, variety and veracity of data – 80% of it unstructured – is growing at a rate impossible to keep up with. Customers have a wider range of choices than ever before and are expecting innovative, relevant and personalized engagement. Why is Cognitive Important? Companies must engage customers on their terms - in a consistent, natural, and intuitive way. Cognitive is the new competitive advantage for enterprises focused on enhancing the customer experience.
- 10. 10 Column Value Patient Joe Brown Date of Birth 02/13/1972 Date Admitted 02/05/2014 Structured Data High Degree of organization, such as a relational database “The patient came in complaining of chest pain, shortness of breath, and lingering headaches…smokes 2 packs a day… family history of heart disease…has been experiencing similar symptoms for the past 12 hours….” Unstructured Data Information that is difficult to organize using traditional mechanisms Structured vs. Unstructured Data
- 11. 11 explorer India In May 1898 India In May celebrated anniversary in Portugal In May, Gary arrived in India after he celebrated his anniversary in Portugal Portugal 400th anniversary celebrated Gary In May 1898 Portugal celebrated the 400th anniversary of this explorer’s arrival in India This evidence suggests “Gary” is the answer BUT the system must learn that keyword matching may be weak relative to other types of evidence arrived in arrival in Legend Keyword “Hit” Reference Text Answer Weak evidenceRed Text Answering complex natural language questions requires more than keyword evidence Analyzing Unstructured Content
- 12. 12 27th May 1498 Vasco da Gama landed in arrival in explorer India Para- phrases Geo- KB Date Match Stronger evidence can be much harder to find and score … … and the evidence is still not 100% certain Search far and wide Explore many hypotheses Find judge evidence Many inference algorithms On the 27th of May 1498, Vasco da Gama landed in Kappad Beach 400th anniversary Portugal May 1898 celebrated In May 1898 Portugal celebrated the 400th anniversary of this explorer’s arrival in India. Kappad Beach Legend Temporal Reasoning Reference Text Answer Statistical Paraphrasing GeoSpatial Reasoning Leverage Multiple Algorithms The Watson Difference:
- 14. 14 Scripted vs. Cognitive Conversations • Driven by a pre-defined conversation flow • Expects key phrases or words • Functions best on structured data • Best for short and simple tasks • Relatively quick to implement Scripted Conversations • Driven by conversational intents rather than expected flow • Trained to understand natural language • Operates on both structured and unstructured data • Learns over time • Capable of a wide range of tasks • Training time varies by complexity Cognitive Conversations
- 15. 15 Conceptual Virtual Agent Architecture
- 16. 16 Virtual Agent Knowledge Base Frequency Complexity High complexity, answer depends on a number of variables (knowing the intent is not enough to answer), requires Deep QA search. Short Tail Long Tail Proof-of-Technology Phase1 Phase2 Phase 3 Low complexity, easy to answer derived using context of the question itself Phase 4 Phase 5
- 17. 17 Forrester Wave Q2 2018 Conversational Computing Platforms
- 19. 19 Virtual Agent Vocabulary I need to add my daughter to my auto policy. utterance entity entity intent: addDriver verb noun
- 20. 20 Intent Training I need to add my daughter to my auto policy. Training Set: intent: addDriver My son just turned 16 and I need to add him to my policy. I have to update my policy to include my nanny. Make sure my account covers my twins also. Test Set Please add my son.
- 21. 21 Entity Training I need to add my daughter to my auto policy. Training Set: entity: PolicyType - auto My son just turned 16 and I need to add him to my policy. What does my automotive insurance cover? My twins are new drivers please add them to my policy. Test Set Does my car insurance cover theft? requires inference
- 22. Watson Assistant
- 23. 23 IBMWatson’sAI capabilities organized into 6 categories and available as API’s on IBM Cloud © 2018 IBM Corporation Knowledge Natural Language Classifier Language Translator Speech Text to SpeechSpeech to Text Empathy Tone Analyzer Personality Insights Language Natural Language Understanding Discovery incl Element Classification Discovery News Vision Visual Recognition Watson Knowledge Studio AI Assistants & Chatbots Watson Assistant Watson sees, hears, speaks, feels, converses, translates, finds http://www.ibm.com/watson/products-services 2
- 24. 24 3 Primary Watson Assistant Use Cases Customer Care Through the Watson Assistant, IBM can decrease call center operations cost, while improving the customer experience and developing new revenue streams Conversational Commerce Provide guided buying experience for prospective customers to purchase goods and services through the mobile or messaging channel of their choice Employee Productivity Simplify access to common questions and tasks through enterprise channels © 2018 IBM Corporation
- 25. 25 Contextual Entities Enable entities to be contextually aware and expand value Recently Released Features • Dialog Folders Stay organized as you scale your bot. • Digressions Dynamically answer questions in the midst of a business process • Rich Response Types Provide buttons, images, videos, pauses, etc. into the response to the end user Set Context in the Dialog UI Avoid having to set context in the advanced JSON editor. Conflict Detection (Premium) Find conflicts between intents and resolve them 6 Entity Synonym Recommendations Disambiguation (Premium) Prompt user for clarification on which intent was intended 7 8 Separate Log Files Separate workspace from your log file, allowing you to improve a bot while in production. 9 Based on a particular value, recommend other synonyms for the user
- 26. Case Studies
- 27. 27 A Digital Concierge Reshaped the User Experience Autonomously Handles Tier-1 Requests (60% Upon Initial Release) Supports Software Activation and Maintenance Tasks 300% Increase in Web Traffic 90% 99% lower support costs shorter resolution times North American Software Provider
- 28. 28 63%reduced AHT Interactive Agent for Healthcare Providers Cognitive Agent Converses with Providers to Verify Benefits Seamlessly Manages Member Information Inquiries Transformed a Tedious IVR System Drastic Reduction in Live Agent Requests Call Time Reduced from 8 to 3 Minutes live agent requests Major Health Insurer
- 29. 29 Bradesco is one of the biggest and fastest growing banking and financial services companies in Brazil. The bank recognized that international expansion could provide some growth opportunities but saw the need to improve service to existing customers as a top priority. Working with Watson, Bradesco, created a virtual customer service solution to provide support for call center agents. The agents, who answered branch agent queries, now have an AI system that recognizes, understands and answers these questions with a high level of accuracy. 283k+ questions is answered by Watson in Portuguese every month 95% Watson is answering questions at a 95% accuracy rate, with only 5% requiring calls for further assistance 62 products Watson has been trained on 62 different Bradesco products use Watson to better serve Bradesco’s 65 million customers 5200 branches 2 © 2018 IBM Corporation
- 30. 30 In an age when elegant technology interactions have become key factors in many consumer decisions, financial institutions must strive to digitally differentiate themselves from the competition. RBS recognized that although its customer service representatives were crucial to a customer’s journey, they spent too much time handling problems that are easily resolved, increasing wait times and negatively affecting satisfaction ratings. RBS used the IBM Watson Conversation Service to build a conversational solution called Cora. The chatbot weaves automated assistance seamlessly with human intervention, creating a hybrid solution that serves as a virtual assistant. Cora has reduced the length of agent/customer conversations by 20% 200+ question topics per brand within Cora’s knowledge base, which has more than doubled since 50% of customer questions Cora can answer after just a few months 85% of customer inquiries are managed by Cora 2x The average contact center agent can now spend up to twice as much time on the more complex problems and questions that matter most to customers 3 © 2018 IBM Corporation
- 31. Questions Type your question into the chat box
- 32. Thank You
Editor's Notes
- #8: Stats from Gartner 2019: By 2019, 20 percent of brands will abandon their mobile apps. Many brands are finding that mobile applications are not delivering the level of adoption and customer engagement they expected. Original return-on-investment (ROI) calculations are missing the mark due to the cost of support, maintenance, upgrades, customer care and marketing to drive downloads. Brands are now investing to build presence in consumer messaging apps, such as Facebook Messenger and WeChat, to reach customers where they spend a high percentage of their time. Source: Gartner Predict 2018: Artificial Intelligence https://www.gartner.com/newsroom/id/3858564 Gartner Says 25 Percent of Customer Service Operations Will Use Virtual Customer Assistants by 2020 Analysts Present Top Predictions for Customer Experience Leaders at the Gartner Customer Experience Summit 2018 in Tokyo, February 19-20
- #12: We faced a lot of technical challenges but at the center of the problem is dealing with the many was you can express the same meaning in natural language. NL is often very sensitive to context and is often incomplete, tacit and ambiguous. Simplified approaches can easily lead you astray. These next two examples should help motivate our approach. Consider this question. <Read it> Now consider that based simply on keywords it would be straight-forward to pick up this potentially answer-bearing passage. <read green passage> This is a great hit from a keyword perspective in shares many common terms – May, Arrived, Anniversary, Portugal, India etc. and by using keyword evidence should give good confidence that Gary is the explorer in question. And whose to say Garry is not an Explorer. After all, we are all explorers in our own special way. In fact, the next sentence might read – and then Gary returned home to explore his attic looking for a lost photo album. Such a sentence would be legitimate evidence that Gary can be classified as an Explorer. Classifications are tricky, we humans are very flexible in how we classify things – we are willing to accept all sorts of variations in meaning to make language work. Of course in this case, the famous explorer Vasco De Gama is the correct answer but how would a computer know that for sure. A computer system must learn to dig deeper, to find, evaluate and weigh different kinds of evidence – ultimately finding the answer that is best supported by the content. Consider this…<next slide>
- #13: Here we see the same question on the right <read it again> To identify and gain confidence in better evidence, the system must parse the question, determining its grammatical structure and identify the main predicates like celebrated and arrived along with their main arguments (that is their subjects and objects, etc) for example -- who is doing the celebrating, and who is doing the arriving AND for each of these actions where and when are they happening. This would further require the system to attempt to distinguish places, dates and people from each other and from other words and phrases in the question. On the right side, we see a passage containing the RIGHT answer BUT with only one key word in common -- “MAY”. <read the green passage> Given just that one common and very popular term, the system must look at a huge amount of unrelated stuff to even get a chance to consider this passage and then must employ and weigh the right algorithms to match the question with an accurate confidence, for example in this case <click> Temporal reasoning algorithms can relate a 400th anniversary in 1898 to 1498, Statistical Paraphrasing algorithms can help the computer learn from reading lots of texts that landed in can imply arrived in and finally with Geospatial reasoning using geographical databases the system may learn that Kappad Beach is in India and if you arrive in Kappad Beach you have therefore arrived in India. And still, all of this will admit numerous errors since few of these computations will produce 100% certainty in mapping from words, to concepts to other words. Just as an example, what if the passage said “considered landing in” rather than “landed in” or what if it the question said “arrival in what he thought to be India?”. Question Answering Technology tries to understand what the user is really asking for and to deliver precise and correct responses. But Natural language is hard … the authors intended meaning can be expressed in so many different ways. To achieve high levels of precision and confidence you must consider much more information and analyze it more deeply. We needed a radically different approach that could rapidly admit and integrate many algorithms, considering lots of different bits of evidence from different perspectives, AND that could learn how to combine and weigh these different sorts of evidence ultimately determining how strongly or weakly they support or refute possible answers.
- #20: https://chatbotsmagazine.com/chatbot-vocabulary-10-chatbot-terms-you-need-to-know-3911b1ef31b4 Utterance: Anything the user says. For example, if a user types “show me yesterday’s financial news”, the entire sentence is the utterance. Intent: An intent is the user’s intention. For example, if a user types “show me yesterday’s financial news”, the user’s intent is to retrieve a list of financial headlines. Intents are given a name, often a verb and a noun, such as “showNews”. Entity: An entity modifies an intent. For example, if a user types “show me yesterday’s financial news”, the entities are “yesterday” and “financial”. Entities are given a name, such as “dateTime” and “newsType”. Entities are sometimes referred to as slots.
- #21: https://chatbotsmagazine.com/chatbot-vocabulary-10-chatbot-terms-you-need-to-know-3911b1ef31b4 Utterance: Anything the user says. For example, if a user types “show me yesterday’s financial news”, the entire sentence is the utterance. Intent: An intent is the user’s intention. For example, if a user types “show me yesterday’s financial news”, the user’s intent is to retrieve a list of financial headlines. Intents are given a name, often a verb and a noun, such as “showNews”. Entity: An entity modifies an intent. For example, if a user types “show me yesterday’s financial news”, the entities are “yesterday” and “financial”. Entities are given a name, such as “dateTime” and “newsType”. Entities are sometimes referred to as slots.
- #22: https://chatbotsmagazine.com/chatbot-vocabulary-10-chatbot-terms-you-need-to-know-3911b1ef31b4 Utterance: Anything the user says. For example, if a user types “show me yesterday’s financial news”, the entire sentence is the utterance. Intent: An intent is the user’s intention. For example, if a user types “show me yesterday’s financial news”, the user’s intent is to retrieve a list of financial headlines. Intents are given a name, often a verb and a noun, such as “showNews”. Entity: An entity modifies an intent. For example, if a user types “show me yesterday’s financial news”, the entities are “yesterday” and “financial”. Entities are given a name, such as “dateTime” and “newsType”. Entities are sometimes referred to as slots.
- #26: new slids