<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Rajeev Shrivastava on Medium]]></title>
        <description><![CDATA[Stories by Rajeev Shrivastava on Medium]]></description>
        <link>https://medium.com/@rshriv?source=rss-6d11dc06c231------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/2*yYgCE5LaekIc2yiBwWnn5Q.jpeg</url>
            <title>Stories by Rajeev Shrivastava on Medium</title>
            <link>https://medium.com/@rshriv?source=rss-6d11dc06c231------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 22:24:17 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@rshriv/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Reference Architecture for Cognitive Contact Center — Part 1]]></title>
            <link>https://medium.com/@rshriv/reference-architecture-for-cognitive-contact-center-part-1-970fabf773a2?source=rss-6d11dc06c231------2</link>
            <guid isPermaLink="false">https://medium.com/p/970fabf773a2</guid>
            <category><![CDATA[contact-center]]></category>
            <category><![CDATA[text-to-speech]]></category>
            <category><![CDATA[voice-agent]]></category>
            <category><![CDATA[aiconversation]]></category>
            <category><![CDATA[natural-language-processi]]></category>
            <dc:creator><![CDATA[Rajeev Shrivastava]]></dc:creator>
            <pubDate>Fri, 31 Jul 2020 23:00:04 GMT</pubDate>
            <atom:updated>2020-07-31T23:00:04.146Z</atom:updated>
            <content:encoded><![CDATA[<h3>Reference Architecture for Cognitive Contact Center — Part 1</h3><p>Today’s digitally-savvy consumers are demanding unprecedented levels of 24x7x365 customer service. But most call centers are slow to respond due to aging technology and processes such as Interactive Voice Response Systems (IVR). Consumers want fast, accurate, and personalized interactions to solve their problems. Companies can bring impactful changes to the Contact Centers by using AI-powered virtual agents. Virtual Assistant leverages various services like Personalization, Next best action, Recommendations, and Guided Resolution to improve customer services. Features like transfer customer call to live agents and advance voice of customer analytics transforming the customer care, while reducing costs and delighting the users.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/812/1*WXC82IupIG4EANqVjioTuw.png" /></figure><h3><strong>Metrics and KPI</strong></h3><p><strong>Quantitative metrics</strong> are measured by numbers or statistics. <strong>Qualitative metrics</strong> are measured by observation or assessment through non-numerical reporting. A qualitative metric may be something as simple as “yes” or “no” feedback. Following table listed top matrices of cognitive contact center:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YxxtyzJ0Tf1RPk3_5_vVAw.png" /><figcaption>Table 1: Matrix and KPI</figcaption></figure><h3>Reference Architecture</h3><p>The Architecture leverages various AI-based capabilities such as Natural Language Understanding, Natural Language Classifier, Speech to Text, Text to Speech, Conversation AI, Insight Engine, to address multiple customer problems. This architecture also utilizes multi &amp; hybrid cloud deployments, back-end systems, and third party services. The architecture components are based on industry best practices supporting the overall digital transformation of any enterprise.</p><p>Customers can contact the Call Center via Omni channels such as <strong>Phone</strong>, <strong>Mobile</strong> <strong>Devices</strong>, <strong>SMS</strong>, <strong>Web Chat</strong>, or <strong>Email</strong>s. Based on a specific channel system can dispatch the call to the respective virtual assistant. The Cognitive Contact Center Architecture, <strong>depicted in Diagram</strong>, provides three popular variations of a following self-service virtual assistant to help customers for their firsthand issues:</p><ul><li><strong>Email Processor Virtual Assistant</strong>: In this virtual assistant, email requests can be first parsed by Natural Language Understanding to identify the purpose and then classify by Natural Language Classifier to identify appropriate departments to forward it for further processing.</li><li><strong>Voice Enabled Virtual Assistant: </strong>In<strong> </strong>this<strong> </strong>virtual<strong> </strong>assistant<strong>, </strong>customer is contacting the call center via phone or mobile devices. Voice Agent is then intercepting the call and invoking Speech to Text to convert it to text and then forwarding the text to Watson Assistant to identify the intent for further processing. After processing, the response can then be converting to speech to return to the customer via voice agent.</li><li><strong>SMS/Text based Virtual Assistant:</strong> This is one of the popular Virtual Assistant. It can process SMS or text-based requests. The agent handles SMS or Text-based request and forward to Conversation Service. It returns the response either as simple text or supporting documents.</li></ul><figure><img alt="Reference Architecture for Cognitive Contact Center" src="https://cdn-images-1.medium.com/max/1024/1*aXRgzDByt2QXeJKHSpusUw.png" /><figcaption>Diagram: Reference Architecture for Cognitive Contact [Call] Center</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jxfxvTds3Q9SqxdD7_IxdA.png" /><figcaption>Table 2: Captions for Reference Architecture Diagram</figcaption></figure><h3>Solution Benefits</h3><p>The cognitive contact center solution will enable customers to minimize the dependency on higher cost channels by infusing Artificial Intelligence [AI] to tackle common customer service questions. This solution also provides the following key benefits to the customer:</p><ul><li>It enables delivery of a better experience to customers, earning their loyalty, and improving retention.</li><li>Saves 10 min per customer interaction</li><li>86% customer satisfaction score, encouraging the customer to return</li><li>Improves the operational performance by reducing no. of calls</li><li>75% reduction in average call time</li><li>Eliminates the call waiting</li><li>Saves approximately $1 per minute</li><li>Improve agent efficiency by providing them the real-time knowledge base access</li><li>75K decrease in average handle time</li><li>Generate new revenue by up-selling and cross-selling new products</li><li>$1M+ Potential cost-saving each year with self-service customer service</li><li>Innovation that provides deep cognitive insights</li></ul><h3>Assets / References / Contacts</h3><ul><li>Implement Virtual Assistant via Watson Assistant: <a href="https://www.ibm.com/cloud/watson-assistant/">https://www.ibm.com/cloud/watson-assistant/</a></li><li>More about Natural Language Understanding: <a href="https://www.ibm.com/watson/services/natural-language-understanding/">https://www.ibm.com/watson/services/natural-language-understanding/</a></li><li>Natural Language Classifier to help in classification of the data: <a href="https://www.ibm.com/watson/services/natural-language-classifier/">https://www.ibm.com/watson/services/natural-language-classifier/</a></li><li>Speech to Text Service to convert Speech to text: <a href="https://www.ibm.com/watson/services/speech-to-text/">https://www.ibm.com/watson/services/speech-to-text/</a></li><li>Text to Speech to convert text to speech: <a href="https://www.ibm.com/watson/services/text-to-speech/">https://www.ibm.com/watson/services/text-to-speech/</a></li><li>Insight Engine that is used to generate customer insight from their data: <a href="https://www.ibm.com/cloud/watson-discovery">https://www.ibm.com/cloud/watson-discovery</a></li><li>Voice gateway to intercept customer request and connect to appropriate AI Services and transfer call to live agents: <a href="https://www.ibm.com/us-en/marketplace/voice-gateway">https://www.ibm.com/us-en/marketplace/voice-gateway</a></li><li>More about Agent Assist solution: <a href="https://www.ibm.com/watson/ai-customer-service/agent-assist">https://www.ibm.com/watson/ai-customer-service/agent-assist</a></li><li>AI for customer care example and benefit: <a href="https://www.ibm.com/watson/ai-customer-service">https://www.ibm.com/watson/ai-customer-service</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=970fabf773a2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Auto AI: A Productive Tool for Data Scientists]]></title>
            <link>https://medium.com/@rshriv/auto-ai-a-productive-tool-for-data-scientists-116e59435b93?source=rss-6d11dc06c231------2</link>
            <guid isPermaLink="false">https://medium.com/p/116e59435b93</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[automation]]></category>
            <category><![CDATA[auto-ai]]></category>
            <category><![CDATA[machine-learning-ai]]></category>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[Rajeev Shrivastava]]></dc:creator>
            <pubDate>Thu, 04 Jun 2020 13:13:46 GMT</pubDate>
            <atom:updated>2020-06-04T13:21:07.902Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/0*7Ue1m6dMhqDC8x6a" /></figure><p>Despite a growing awareness of the strategic value of AI in business, most organizations still struggle with fundamental information architecture. The chores of finding, collecting, organizing siloed data, preparing that data for analysis, and then infusing AI are often slowing AI development tasks.</p><p>As per the 2020 AI prediction, an enterprise needs to produce and consume AI models quickly and easily. To achieve this goal, automation will play an important role, especially in the area of AI.</p><p>Auto AI is a powerful set of capability in <a href="https://c212.net/c/link/?t=0&amp;l=en&amp;o=2493908-1&amp;h=2426202198&amp;u=https%3A%2F%2Fwww.ibm.com%2Fcloud%2Fwatson-studio&amp;a=Watson+Studio">Watson Studio</a> to automate many of the often complicated and laborious tasks associated with designing, optimizing, and governing AI models in the enterprise. As a result, data scientists can be freed up to dedicate more time to designing, testing, and deploying machine learning (ML) models.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ItfGtlZMOTvG9gXV_WlaHA.png" /><figcaption>Watson AUTO AI</figcaption></figure><p>One of our customers was able to produce [end-to-end] ten machine learning models in two hours to create business insight. They were quite amazed to see this powerful feature of Watson Studio.</p><h3><strong>Top Challenges:</strong></h3><p>Here are the top challenges that customer is facing in the area of data science and machine learning world:</p><ul><li>80% of data scientists’ time is spent on tedious tasks that could be automated</li><li>63% of companies see the availability of technical skills as a challenge to implementation</li><li>70% of enterprises will boost employee productivity with AI by 2021</li><li>60% of companies see regulatory constraints as a barrier to implementing AI</li><li>Scientists will grow 5X faster 5x than expert data scientists</li></ul><h3><strong>Why Auto AI?</strong></h3><p>Auto AI addresses the above challenges by using the following features:</p><ul><li><strong>Machine Learning Algorithm selection: </strong>It can automatically select the ML algorithm based on your data. The user also can change the auto-detected algorithm by Auto AI<strong>.</strong></li><li><strong>Speed Model Selection: </strong>It can shortlist the top-performing model in minutes instead of days/weeks. It also drastically reduces the neural network search time.</li><li><strong>Jump the Skills Gap: </strong>Citizen data scientists can learn, understand, and analyze the implementation by looking at various input, output, and generated code.</li><li><strong>Drive Productivity: </strong>Citizen data scientists can start the AI experiment without knowing how to code. They can do more innovative work instead of mundane tasks (e.g. lengthy feature selection process)</li><li><strong>One-click deployment: </strong>It<strong> </strong>automates AI life-cycle management. Data scientists can deploy high ranking models by one-click buttons.</li><li><strong>Govern your AI: </strong>Increase trust and transparency by integrating IBM Watson OpenScale. Data scientists can also Monitor AI outcomes with explainability and debiasing</li></ul><h3><strong>What Auto AI automates?</strong></h3><p>Auto AI automated following features to expedite the development and deployment of machine learning models:</p><ul><li><strong>Data Preparation: </strong>It applies various algorithms, or estimators, to analyze, clean, and prepare your raw data for machine learning. It automatically detects and categorizes features based on data type, such as categorical or numerical.</li><li><strong>Model Selection: </strong>It uses a novel approach that enables testing and ranking candidate algorithms against small subsets of the data, gradually increasing the size of the subset for the most promising algorithms to arrive at the best match.</li><li><strong>Feature engineering: </strong>It automates feature engineering that attempts to transform the raw data into the combination of features that best represents the problem to achieve the most accurate prediction.</li><li><strong>Hyper-parameter optimization: </strong>It refines the best performing model pipelines. It uses the hyper-parameter algorithm to optimize for costly function evaluations such as model training and scoring.</li></ul><h4><strong>How does it work?</strong></h4><p>Auto AI uses the following process to build and deploy a machine learning model without writing a single line of code.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3TuGDA9QaFKpkuLlXsf5Vw.png" /><figcaption>Detailed Flow of Auto AI</figcaption></figure><h4><strong>What tools are required to experiment with Auto AI?</strong></h4><p>Following tools or services are required to implement Auto AI in any AI Development:</p><ul><li><strong>Watson Studio</strong> [Local or Cloud version] — The integrated environment to design, develop, deploy, and test models via Auto AI feature.</li><li><strong>Watson Machine Learning </strong>— This service is required to deploy the ML Model.</li><li><strong>Watson Open Scale [Optional]</strong> — This service is required to govern the ML Model in production.</li><li><strong>Object Storage [Optional]</strong> — This service is required to store the data set.</li></ul><h4><strong>Learning Resources:</strong></h4><p>Listen from experts on Need for Automation of AI:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FBwPEmDM1uTA%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DBwPEmDM1uTA&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FBwPEmDM1uTA%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/fcc0ae6d457df7d0bdfb6cb2be7cf7e1/href">https://medium.com/media/fcc0ae6d457df7d0bdfb6cb2be7cf7e1/href</a></iframe><p>Interested in getting started? Check out our Auto AI <a href="https://dataplatform.cloud.ibm.com/docs/content/wsj/analyze-data/autoai_example_binary_classifier.html">Tutorial</a> and Data <a href="https://dataplatform.cloud.ibm.com/docs/content/wsj/analyze-data/wml-ai.html">Documentation</a> to start learning!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=116e59435b93" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Watson Assistant Search Skill: Taking Virtual Assistant to the next level]]></title>
            <link>https://medium.com/ibm-watson/watson-assistant-search-skill-taking-virtual-assistant-to-the-next-level-3c122aa8a110?source=rss-6d11dc06c231------2</link>
            <guid isPermaLink="false">https://medium.com/p/3c122aa8a110</guid>
            <category><![CDATA[wa-tutorial]]></category>
            <category><![CDATA[insight-engine]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[virtual-assistant]]></category>
            <category><![CDATA[watson-assistant]]></category>
            <dc:creator><![CDATA[Rajeev Shrivastava]]></dc:creator>
            <pubDate>Sun, 24 May 2020 02:40:00 GMT</pubDate>
            <atom:updated>2020-06-17T19:46:00.886Z</atom:updated>
            <content:encoded><![CDATA[<h3>Watson Assistant Search Skill: Taking Virtual Assistant To The Next Level</h3><p>As the world reacts to the Covid-19 coronavirus pandemic, call centers of hospitals, government agencies, and other organizations are overwhelmed with the number of calls to find out repeated common information. Everyone wants personal interaction to get information from authentic sources. Remote working and collaboration tools have become essential systems, with new demands placed on networks and data center infrastructure. In this situation, virtual assistants are playing vital roles to share the load from human agents and subject matter experts so that they can focus on real business tasks in hand. As a result, the usage of IBM Watson Assistant to implement these virtual assistants is increased by 60%. The Dialog skill and Search skills of the IBM Watson Assistant are essential components in these implementations. The next couple of sections of this article will provide you with information on how the Search skill of Watson Assistant is enabling virtual assistants or bots to rise up to the next level of human interaction.</p><h4><strong>What is Search Skill?</strong></h4><p>When Watson Assistant doesn’t have an explicit solution to a problem, it routes the user’s question to a search skill to find an answer from across your disparate sources of self-service content. The search skill interacts with the IBM Watson™ Discovery service to extract this information from a configured data collection. For example, a remote worker is asking questions such as, “How to setup a VPN?” or, “What are the steps to access a shared network drive?” Answers to these questions sometimes require a document that describes a number of steps.</p><p>The following diagram illustrates another example to show end-to-end integration flow of search skill between Watson Assistant and Watson Discovery:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/936/1*bYTdRFIRtDbl3gZN7pzolQ.png" /><figcaption>End-to-End Integration Flow for</figcaption></figure><h4><strong>When you need Search Skill?</strong></h4><ul><li><strong>Dynamic situation:</strong> If you want to take your bot or virtual assistant to the next level as close to human interaction then you need to use search skill to handle the dynamic situation like Covid-19, where information is changing very fast based on new facts. Search skill can help you in this situation by fetching the document crawl from the website.</li><li><strong>For Longtail Queries:</strong> If you want to present the response as a full document or paragraph within it. For example, the query<em> What are the benefits available for full-time employees?</em> Response for this query requires a full document or paragraph related to employee benefits.</li><li><strong>Low Confidence:</strong> If your virtual assistant or bot has low confidence in a specific query to respond, it can present a number of documents from the knowledge base.</li></ul><h4><strong>How to Trigger Search Skill?</strong></h4><p>Search Skill can be triggered in the following ways:</p><ul><li><strong>From Anything Else Node</strong>: Searches an external data source for a relevant answer when none of the dialog nodes can address the user’s query.</li><li><strong>From Dialog Node</strong>: If you add a search skill response type to a dialog node, then your bot retrieves documents or a passage from an external data source as the response to a particular question. This approach is useful if you want to narrow down a user query before you trigger a search.</li><li><strong>For Low Confidence Score: </strong>Search skills can also be triggered based on the confidence score. Add a folder to the dialog. Position the folder above the first dialog node that you want to de-emphasize. Add the following condition to the folder: <strong><em>intents[0].confidence &lt; 0.5</em></strong></li><li><strong>Search Skill Only</strong>: If only a search skill is linked to an assistant, and no dialog skill is present, then a search query is sent to the Discovery service when any user input is received from one of the assistant’s integration channels.</li></ul><h4><strong>How to Configure Search Skills?</strong></h4><p>For configuring Search skill, you can follow the detailed video to configure search skill:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F_UgRPaxipgI%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D_UgRPaxipgI&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2F_UgRPaxipgI%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/d8cd6d373e84d149bcbc8b30d20f65ce/href">https://medium.com/media/d8cd6d373e84d149bcbc8b30d20f65ce/href</a></iframe><h4><strong>How to optimize the Search?</strong></h4><p>Most of the time, discovery service is returning the desirable documents to the search skill. However, there are ways you can improve or optimize the search results in Watson Discovery Services:</p><ul><li><strong>Search settings</strong>:There are two techniques you can use to optimize search: 1. <strong><em>Query Expansion:</em></strong> You can expand the scope of a query beyond exact matches — for example, you can expand a query for “Covid-19” to include “Coronavirus” and “Novelcoronavirus “ — by uploading a list of query expansion terms. Query expansion terms are usually synonyms, antonyms, or typical misspellings for common terms. For more information, see this link: <a href="https://cloud.ibm.com/docs/discovery?topic=discovery-query-concepts#stopwords"><em>Query Expansion in Discovery</em></a> 2. <strong><em>Defining Stopwords</em></strong>: Stopwords are words that are filtered out of queries because they add little value, for example: a, an, the. Adding common words to a stopwords list can also improve the relevance of results for natural language queries. Discovery applies a default list of stopwords for several languages at query time. However, you can define and upload a custom list of stopwords that override the default list. For more information, see this link: <a href="https://cloud.ibm.com/docs/discovery?topic=discovery-query-concepts#stopwords"><em>Defining Stopwords</em></a></li><li><strong>Smart Document Understanding [SDU]</strong>: Sometimes, you want to show an appropriate section in the document as a response to the search query. By SDU, you can control the nuance in the document structure. This feature is pretty useful while using search skill. For more information, see the link: <a href="https://cloud.ibm.com/docs/discovery?topic=discovery-sdu"><em>Smart Document Understanding</em></a></li><li><strong>Relevancy Training</strong>: If nothing works and you still need improvement in search results, then try relevancy training. Relevancy training is optional; if the results of your queries meet your needs, no further training is necessary. The relevance of natural language query results can be improved in IBM Watson™ Discovery with training. You can train your private collections using either the Discovery tooling or the Discovery APIs. For more information, see the link to apply relevancy training in your collection: <a href="https://cloud.ibm.com/docs/discovery?topic=discovery-improving-result-relevance-with-the-tooling"><em>Relevancy Training for Search Results</em></a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3c122aa8a110" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ibm-watson/watson-assistant-search-skill-taking-virtual-assistant-to-the-next-level-3c122aa8a110">Watson Assistant Search Skill: Taking Virtual Assistant to the next level</a> was originally published in <a href="https://medium.com/ibm-watson">IBM watsonx Assistant</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>