{"id":55911,"date":"2023-03-15T09:05:15","date_gmt":"2023-03-15T09:05:15","guid":{"rendered":"https:\/\/developer.kore.ai\/v10-1\/?p=55911"},"modified":"2024-02-22T08:15:09","modified_gmt":"2024-02-22T08:15:09","slug":"llm-and-generative-ai","status":"publish","type":"post","link":"https:\/\/developer.kore.ai\/v10-1\/docs\/bots\/nlp\/llm-and-generative-ai\/","title":{"rendered":"Introduction to LLM and Generative AI"},"content":{"rendered":"<section class=\"l-section wpb_row height_auto\"><div class=\"l-section-h i-cf\"><div class=\"g-cols vc_row type_default valign_top\"><div class=\"vc_col-sm-12 wpb_column vc_column_container\"><div class=\"vc_column-inner\"><div class=\"wpb_wrapper\"><div class=\"ult-spacer spacer-6ae865ca3e87f\" data-id=\"6ae865ca3e87f\" data-height=\"10\" data-height-mobile=\"10\" data-height-tab=\"10\" data-height-tab-portrait=\"\" data-height-mobile-landscape=\"\" style=\"clear:both;display:block;\"><\/div>\r\n\t<div class=\"wpb_text_column \">\r\n\t\t<div class=\"wpb_wrapper\">\r\n\t\t\t<p><span style=\"font-weight: 400;\">The Kore.ai XO Platform helps enhance your bot development process and enrich end-user conversational experiences by integrating pre-built (OpenAI, Azure OpenAI, Anthropic) or custom models in the backend.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">In addition to the out-of-box integration with pre-built models, the Platform supports the bring-your-own (BYO) model framework to integrate with externally hosted models by third parties as well as models hosted by the enterprises themselves. The framework allows the creation of custom prompts that are optimized for specific purposes and models. This generic framework works seamlessly with the Auth Profiles module of the Platform, enabling enterprises to use the authentication mechanism of their choice.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">The new Kore.ai XO GPT Models module provides fine-tuned large language models optimized for enterprise conversational AI applications. These models have been evaluated and fine-tuned to be accurate, safe, and efficient for production deployment. For more information, see <a href=\"https:\/\/developer.kore.ai\/v10-1\/docs\/bots\/nlp\/kore-xo-gpt-model\/\" target=\"_blank\" rel=\"noopener noreferrer\">Kore.ai XO GPT<\/a>.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">By leveraging LLM and Generative AI capabilities, you can create intelligent, human-like conversational experiences for your end-users. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">You can find Generative AI and LLM features by going to <\/span><b>Build<\/b><span style=\"font-weight: 400;\"> &gt; <\/span><b>Natural Language<\/b><span style=\"font-weight: 400;\"> &gt; <\/span><b>Generative AI &amp; LLM<\/b><span style=\"font-weight: 400;\">.<\/span><a ref=\"magnificPopup\" href=\"https:\/\/kore-wordpress.s3.us-east-2.amazonaws.com\/developer.kore.ai\/v10-1\/wp-content\/uploads\/sites\/35\/20240222081239\/llm-intro.png\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone wp-image-61295 size-full\" src=\"https:\/\/kore-wordpress.s3.us-east-2.amazonaws.com\/developer.kore.ai\/v10-1\/wp-content\/uploads\/sites\/35\/20240222081239\/llm-intro.png\" alt=\"\" width=\"1915\" height=\"823\" srcset=\"https:\/\/kore-wordpress.s3.us-east-2.amazonaws.com\/developer.kore.ai\/v10-1\/wp-content\/uploads\/sites\/35\/20240222081239\/llm-intro.png 1915w, https:\/\/kore-wordpress.s3.us-east-2.amazonaws.com\/developer.kore.ai\/v10-1\/wp-content\/uploads\/sites\/35\/20240222081239\/llm-intro-300x129.png 300w, https:\/\/kore-wordpress.s3.us-east-2.amazonaws.com\/developer.kore.ai\/v10-1\/wp-content\/uploads\/sites\/35\/20240222081239\/llm-intro-1024x440.png 1024w\" sizes=\"auto, (max-width: 1915px) 100vw, 1915px\" \/><\/a><\/p>\n<h2><span class=\"ez-toc-section\" id=\"Key_Features\"><span style=\"font-weight: 400;\">Key Features<\/span><\/span><\/h2>\n<p><span style=\"font-weight: 400;\">The Integration of Generative AI and LLM enables the following features:<\/span><\/p>\n<ul>\n<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\">Prompts &amp; Requests Library<\/span><span style=\"font-weight: 400;\">: <\/span><span style=\"font-weight: 400;\">Complete flexibility to create fully customized prompts optimized for particular use cases using custom models.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\">Co-Pilot features:<\/span>\n<ul>\n<li style=\"font-weight: 400;\" aria-level=\"2\"><strong>Automatic Dialog Generation<\/strong><span style=\"font-weight: 400;\">: This feature helps build production-ready dialog tasks automatically by briefly describing the task. A preview of the generated dialog is available and lets you modify the intent description and create multiple iterations of the dialog.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"2\"><strong>Conversation Test Cases Suggestion<\/strong><span style=\"font-weight: 400;\">: The Platform suggests simulated user inputs covering various scenarios from an end-user perspective at every test step. You can use these suggestions to create test suites.<\/span><\/li>\n<li><span style=\"font-weight: 400;\"><strong>Conversation Summary<\/strong>: Implements the Conversation Summary public API to fetch the details of the entire conversation between the customer and the VA or agent. This API leverages the Flan-T5 foundational model to summarize conversations.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"2\"><strong>NLP Batch Test Cases Suggestion<\/strong><span style=\"font-weight: 400;\">: The Platform generates NLP test cases for every intent, including entity checks. You only need to create test suites in the Builder using the generated testing utterances.<\/span><\/li>\n<li><strong>Training Utterance Suggestions<\/strong><span style=\"font-weight: 400;\">: Generate high-quality training data quickly and easily with our platform\u2019s suggested utterances for each intent. Review and add the suggestions as needed to create a powerful training set for your bot.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"2\"><span style=\"font-weight: 400;\"><strong>Use Case Suggestions<\/strong>: Uses the Open AI LLM model to generate use cases during the VA (Bot) creation journey.<\/span><\/li>\n<\/ul>\n<\/li>\n<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\">Dynamic Conversations features:<\/span>\n<ul>\n<li style=\"font-weight: 400;\" aria-level=\"2\"><strong>GenAI Node<\/strong><span style=\"font-weight: 400;\">: Collect Entities from end-users in a free-flowing conversation using LLM and Generative AI in the background. You can define entities to be collected as well as rules &amp; scenarios.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"2\"><strong>Answer from Documents<\/strong><span style=\"font-weight: 400;\">: Helps answer end-user queries from unstructured PDF documents without the need to extract individual FAQs and train them.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"2\"><strong>GenAI Prompt<\/strong><span style=\"font-weight: 400;\"><strong>:<\/strong> Leverage this node to unlock the power of Generative AI with your prompts, enabling you to build creative and custom use cases.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"2\"><strong>Rephrase Dialog Responses<\/strong><span style=\"font-weight: 400;\">: Enhance end-user experience with empathetic and contextual bot responses.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"2\"><strong>Zero-shot ML Model<\/strong><span style=\"font-weight: 400;\">: Uses the Open AI LLM model for intent identification during run time based on semantic similarity.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"2\"><strong>Repeat Responses<\/strong><span style=\"font-weight: 400;\">: Use LLM to reiterate the last bot responses when the Repeat Bot Response event is triggered.<\/span><\/li>\n<li aria-level=\"2\"><span style=\"font-weight: 400;\"><strong>Rephrase User Query<\/strong>: Improve intent detection and entity extraction by enriching the user query with relevant details from the ongoing conversation context.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"2\"><strong>Few-shot ML Model<\/strong><span style=\"font-weight: 400;\">: Uses the Kore Ai\u2019s hosted embeddings for intent identification during run time based on semantic similarity.<\/span><\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<h2><span class=\"ez-toc-section\" id=\"Benefits\"><span style=\"font-weight: 400;\">Benefits<\/span><\/span><\/h2>\n<p><span style=\"font-weight: 400;\">All these features benefit VA developers, NLP developers, and testers as follows:<\/span><\/p>\n<ul>\n<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\">Being able to select among Kore.ai XO GPT or custom or pre-built LLM integrations.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\">Developers can create dialog tasks on the fly through the prebuilt Dialog Tasks Flow.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\">Developers can unlock the power of Generative AI with their prompts, enabling them to build creative and custom use cases.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\">Mundane tasks like generating dialog tasks or training utterances are automated to help developers be more productive and focus on other important tasks like enhancing conversation design, creating complex test cases, and more.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\">Testers can ensure that their intent descriptions are meaningful in the right context to generate the right content.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\">The Platform provides suggestions and nudges developers in the right direction for the better design and development of Virtual Assistants. For example, it offers curated use case suggestions while <\/span><a href=\"https:\/\/developer.kore.ai\/v10-1\/docs\/bots\/chatbot-overview\/getting-started-bots\/#Create_a_Standard_Virtual_Assistant\"><span style=\"font-weight: 400;\">creating the VA<\/span><\/a><span style=\"font-weight: 400;\">, including probable user inputs (simulating end-user behavior) in <\/span><a href=\"https:\/\/developer.kore.ai\/v10-1\/docs\/bots\/test-your-bot\/conversation-testing-landing-page\/\"><span style=\"font-weight: 400;\">Conversation Testing<\/span><\/a><span style=\"font-weight: 400;\">. This way, the VA can simulate the end user\u2019s behavior at every conversation step and respond more realistically by considering error scenarios, digressions, and contextual changes.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\">Allows enterprises and advanced users to bypass the platform&#8217;s pre-built integrations. They can establish customized connections to large language models and optimize prompts and requests for their specific use cases.<\/span><\/li>\n<\/ul>\n<h2><span class=\"ez-toc-section\" id=\"Important_Considerations\"><span style=\"font-weight: 400;\">Important Considerations<\/span><\/span><\/h2>\n<p><span style=\"font-weight: 400;\">Generative AI features are available for English and non-English NLU and VA languages on the Kore.ai XO Platform. However, custom LLM-specific features are currently limited to English. To learn more about managing languages for VAs, click <a href=\"https:\/\/developer.kore.ai\/v10-1\/docs\/bots\/advanced-topics\/multi-lingual\/managing-languages-for-multilingual-vas\/\" target=\"_blank\" rel=\"noopener noreferrer\">here<\/a>.<\/span><\/p>\n<p>LLM and Generative AI also require sharing data with third parties: OpenAI (when using the OpenAI integration) or OpenAI and Microsoft (when using the Azure integration).<\/p>\n<h2><span class=\"ez-toc-section\" id=\"Next_Steps\"><span style=\"font-weight: 400;\">Next Steps<\/span><\/span><\/h2>\n<ol>\n<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\"><a href=\"https:\/\/developer.kore.ai\/v10-1\/docs\/bots\/nlp\/llm-integration\/\">Integrate a pre-built or custom or Kore.ai XO GPT LLM<\/a>.\u00a0<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\">(Optional &#8211; only for custom LLM model) <a href=\"https:\/\/developer.kore.ai\/v10-1\/docs\/bots\/nlp\/prompts-and-requests-library\/\">Add Prompts<\/a>.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"1\"><span style=\"font-weight: 400;\">Enable <a href=\"https:\/\/developer.kore.ai\/v10-1\/docs\/bots\/nlp\/co-pilot-features\/\">Co-Pilot<\/a> and <a href=\"https:\/\/developer.kore.ai\/v10-1\/docs\/bots\/nlp\/dynamic-conversations-features\/\">Dynamic Conversations<\/a> features.<\/span><\/li>\n<\/ol>\n\r\n\t\t<\/div>\r\n\t<\/div>\r\n<div class=\"ult-spacer spacer-6ae865ca3ee32\" data-id=\"6ae865ca3ee32\" data-height=\"10\" data-height-mobile=\"10\" data-height-tab=\"10\" data-height-tab-portrait=\"\" data-height-mobile-landscape=\"\" style=\"clear:both;display:block;\"><\/div><\/div><\/div><\/div><\/div><\/div><\/section><section class=\"l-section wpb_row height_auto\"><div class=\"l-section-h i-cf\"><div class=\"g-cols vc_row type_default valign_top\"><div class=\"vc_col-sm-12 wpb_column vc_column_container\"><div class=\"vc_column-inner\"><div class=\"wpb_wrapper\"><div class=\"ult-spacer spacer-6ae865ca3f8e8\" data-id=\"6ae865ca3f8e8\" data-height=\"10\" data-height-mobile=\"10\" data-height-tab=\"10\" data-height-tab-portrait=\"\" data-height-mobile-landscape=\"\" style=\"clear:both;display:block;\"><\/div><\/div><\/div><\/div><\/div><\/div><\/section>\n","protected":false},"excerpt":{"rendered":"The Kore.ai XO Platform helps enhance your bot development process and enrich end-user conversational experiences by integrating pre-built (OpenAI, Azure OpenAI, Anthropic) or custom models in the backend. In addition to the out-of-box integration with pre-built models, the Platform supports the bring-your-own (BYO) model framework to integrate with externally hosted models by third parties as...","protected":false},"author":27,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[203,6],"tags":[],"class_list":["post-55911","post","type-post","status-publish","format-standard","hentry","category-llm-and-generative-ai","category-nlp"],"_links":{"self":[{"href":"https:\/\/developer.kore.ai\/v10-1\/wp-json\/wp\/v2\/posts\/55911","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/developer.kore.ai\/v10-1\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/developer.kore.ai\/v10-1\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/developer.kore.ai\/v10-1\/wp-json\/wp\/v2\/users\/27"}],"replies":[{"embeddable":true,"href":"https:\/\/developer.kore.ai\/v10-1\/wp-json\/wp\/v2\/comments?post=55911"}],"version-history":[{"count":149,"href":"https:\/\/developer.kore.ai\/v10-1\/wp-json\/wp\/v2\/posts\/55911\/revisions"}],"predecessor-version":[{"id":61298,"href":"https:\/\/developer.kore.ai\/v10-1\/wp-json\/wp\/v2\/posts\/55911\/revisions\/61298"}],"wp:attachment":[{"href":"https:\/\/developer.kore.ai\/v10-1\/wp-json\/wp\/v2\/media?parent=55911"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/developer.kore.ai\/v10-1\/wp-json\/wp\/v2\/categories?post=55911"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/developer.kore.ai\/v10-1\/wp-json\/wp\/v2\/tags?post=55911"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}