{"_id":"5845a4a89f6fbb1b004307c4","__v":4,"project":"54d3007669578e0d002730c9","version":{"_id":"5845a4a89f6fbb1b004307b7","project":"54d3007669578e0d002730c9","__v":2,"createdAt":"2016-12-05T17:32:24.708Z","releaseDate":"2016-12-05T17:32:24.708Z","categories":["5845a4a89f6fbb1b004307b8","5845a4a89f6fbb1b004307b9","5845a4a89f6fbb1b004307ba","5845a4a89f6fbb1b004307bb","5845a4a89f6fbb1b004307bc","5845a4a89f6fbb1b004307bd","5845a4a89f6fbb1b004307be","5845a4a89f6fbb1b004307bf","5845a4a89f6fbb1b004307c0","592deb23644f060f008e5aa6"],"is_deprecated":false,"is_hidden":false,"is_beta":false,"is_stable":true,"codename":"","version_clean":"25.0.0","version":"25"},"user":"55a04a395730f40d001104d2","parentDoc":null,"category":{"_id":"5845a4a89f6fbb1b004307b8","__v":0,"version":"5845a4a89f6fbb1b004307b7","project":"54d3007669578e0d002730c9","sync":{"url":"","isSync":false},"reference":false,"createdAt":"2015-02-05T05:32:38.958Z","from_sync":false,"order":0,"slug":"overview","title":"Overview"},"updates":["56b1b3ede026e417001aa636","56bdc5ab5ed4720d005c2b5d","56c733b0dd0b310d00e82ef5","574fdeb78b8fc60e00937b5a","57a8b73d2ad1150e009dab06","585d90f5b24f62390000882a","585d9150e8f8a72f00d57457","5884953374008a2300da6d95","5884953b3b8cd02d00d758ab"],"next":{"pages":[],"description":""},"createdAt":"2015-08-11T22:45:27.651Z","link_external":false,"link_url":"","githubsync":"","sync_unique":"","hidden":false,"api":{"results":{"codes":[]},"settings":"","auth":"required","params":[],"url":""},"isReference":false,"order":1,"body":"* [Step 1: Create agent](#step-1-create-agent)\n* [Step 2: Create entities](#step-2-create-entities)\n* [Step 3: Create intents](#step-3-create-intents)\n* [Step 4: Test and train your agent](#step-4-test-and-train-your-agent)\n* [Step 5: Integrate](#step-5-integrate)\n\n[block:api-header]\n{\n  \"type\": \"basic\"\n}\n[/block]\nOur goal is to make the process of creating and integrating sophisticated conversational interfaces as simple as possible.\n\nIn this Quick Start, we will lead you through a step-by-step guide of how to integrate your product to processing natural language using the API.AI tools.\n[block:api-header]\n{\n  \"type\": \"basic\",\n  \"title\": \"Step 1: Create agent\"\n}\n[/block]\nAn API.AI agent represents a conversational interface of your application, device, or bot.\n\nYou'll be prompted to create an agent right after you register at API.AI. If you are registered already, just <a href=\"https://console.api.ai/api-client/#/newAgent\" target=\"_blank\">follow this link</a> to create a new agent.\n[block:image]\n{\n  \"images\": [\n    {\n      \"image\": [\n        \"https://files.readme.io/7a68a45-getting-started-create-agent.png\",\n        \"getting-started-create-agent.png\",\n        2134,\n        654,\n        \"#3482ac\"\n      ],\n      \"sizing\": \"full\"\n    }\n  ]\n}\n[/block]\nAfter entering the name of your new agent, choose the language you want your app to communicate in. Currently, API.AI supports <a href=\"https://docs.api.ai/docs/languages\" target=\"_blank\">15 languages</a>.\n[block:callout]\n{\n  \"type\": \"warning\",\n  \"body\": \"The language you choose for your agent cannot be changed after saving. However, you can always create a new agent if you wish to work in an additional language.\"\n}\n[/block]\n\n[block:image]\n{\n  \"images\": [\n    {\n      \"image\": [\n        \"https://files.readme.io/c47e862-getting-started-select-lang.png\",\n        \"getting-started-select-lang.png\",\n        2090,\n        710,\n        \"#fafafa\"\n      ],\n      \"sizing\": \"full\"\n    }\n  ]\n}\n[/block]\nOnce you have named your agent and selected a language, click the 'Save' button. \n\nFor more detailed information about agents, read <a href=\"https://docs.api.ai/docs/concept-agents\" target=\"_blank\">this documentation</a>. \n[block:api-header]\n{\n  \"type\": \"basic\",\n  \"title\": \"Step 2: Create entities\"\n}\n[/block]\nOnce you’ve created your agent, you may want to create a few entities to extract necessary information from user queries. These will be stored in the <a href=\"https://console.api.ai/api-client/#/entities\" target=\"_blank\">Entities panel</a> located on the left of the API.AI developer console.\n[block:image]\n{\n  \"images\": [\n    {\n      \"image\": [\n        \"https://files.readme.io/6932fb3-getting-started-entities.png\",\n        \"getting-started-entities.png\",\n        2058,\n        714,\n        \"#2a8ab9\"\n      ],\n      \"sizing\": \"full\"\n    }\n  ]\n}\n[/block]\nEntities represent concepts or objects that are often specific to a use case. They are used for storing possible parameter values (along with their synonyms) that need to be returned in the <a href=\"https://docs.api.ai/docs/query#response\" target=\"_blank\">JSON response to a query</a>.\n\nThe following example from a pizza delivery agent shows the entity titled “type”. It represents different types of pizza that the user can order. One pizza type can have several name variations. All such variations (synonyms) are mapped to a reference value that will be returned in the JSON response (unless other parameter value type is defined).\n[block:image]\n{\n  \"images\": [\n    {\n      \"image\": [\n        \"https://files.readme.io/7d35454-getting-started-entity-example.png\",\n        \"getting-started-entity-example.png\",\n        2048,\n        1060,\n        \"#2e8ab3\"\n      ],\n      \"sizing\": \"full\"\n    }\n  ]\n}\n[/block]\nRead more about <a href=\"https://docs.api.ai/docs/concept-entities\" target=\"_blank\">entities</a> and <a href=\"https://docs.api.ai/docs/concept-actions\" target=\"_blank\">parameters</a>.\n[block:api-header]\n{\n  \"type\": \"basic\",\n  \"title\": \"Step 3: Create intents\"\n}\n[/block]\nOnce you created one or more entities, proceed to the <a href=\"https://console.api.ai/api-client/#/intents\" target=\"_blank\">Intents panel</a>.\n\nAn intent determines what kind of user requests should be understood and to what actionable data (<a href=\"https://docs.api.ai/docs/query#response\" target=\"_blank\">JSON response</a>) those requests should be converted.\n\nBelow is an example intent from a pizza delivery app, which is designed to understand some basic requests from users when they want to start their order.\n[block:image]\n{\n  \"images\": [\n    {\n      \"image\": [\n        \"https://files.readme.io/fafb609-Getting-started_intent.png\",\n        \"Getting-started_intent.png\",\n        2044,\n        1044,\n        \"#f6f5f7\"\n      ],\n      \"sizing\": \"full\"\n    }\n  ]\n}\n[/block]\nNote that the <a href=\"https://docs.api.ai/docs/machine-learning-settings\" target=\"_blank\">Machine Learning</a> option is enabled in all intents by default, which means that your agent will use request examples from the intent to train itself to understand many other variations of similar inputs.\n\nThe more of the relevant and diverse examples you add to the intents, the smarter your agent will be.\n\nRead more about intents <a href=\"https://docs.api.ai/docs/concept-intents\" target=\"_blank\">here</a>.\n[block:api-header]\n{\n  \"type\": \"basic\",\n  \"title\": \"Step 4: Test and train your agent\"\n}\n[/block]\nIn the top right corner of the API.AI developer console, you can find the test console, where you can test how well your agent understands you.\n[block:image]\n{\n  \"images\": [\n    {\n      \"image\": [\n        \"https://files.readme.io/6146708-getting-started-test.png\",\n        \"getting-started-test.png\",\n        2790,\n        714,\n        \"#2894ad\"\n      ],\n      \"sizing\": \"full\"\n    }\n  ]\n}\n[/block]\nIf your request is recognized, you'll see which intent was used to process it and what information was extracted.\n[block:image]\n{\n  \"images\": [\n    {\n      \"image\": [\n        \"https://files.readme.io/0e91ce6-getting-started-testing-results.png\",\n        \"getting-started-testing-results.png\",\n        554,\n        1248,\n        \"#f4f5f5\"\n      ]\n    }\n  ]\n}\n[/block]\nIf your request was not matched to a desired intent or was matched incorrectly, you can add it to one of your intents right away or use the Training tool later to find all the unmatched inputs and use them to train your agent. You can find more information about Training <a href=\"https://docs.api.ai/docs/training\" target=\"_blank\">here</a>. \n[block:api-header]\n{\n  \"type\": \"basic\",\n  \"title\": \"Step 5: Integrate\"\n}\n[/block]\nOnce your agent is ready, integrate it into your app or device. To facilitate this process, we created a number of useful tools:\n\n* <a href=\"https://docs.api.ai/docs/sdks\" target=\"_blank\">SDK's</a> for the most popular languages and platforms\n* <a href=\"https://docs.api.ai/docs/integrations\" target=\"_blank\">One-click integrations</a> for the most popular messaging platforms\n* <a href=\"https://docs.api.ai/docs/webhook\" target=\"_blank\">Webhook</a> integration with your web service","excerpt":"","slug":"get-started","type":"basic","title":"Get started in 5 steps"}

Get started in 5 steps


* [Step 1: Create agent](#step-1-create-agent) * [Step 2: Create entities](#step-2-create-entities) * [Step 3: Create intents](#step-3-create-intents) * [Step 4: Test and train your agent](#step-4-test-and-train-your-agent) * [Step 5: Integrate](#step-5-integrate) [block:api-header] { "type": "basic" } [/block] Our goal is to make the process of creating and integrating sophisticated conversational interfaces as simple as possible. In this Quick Start, we will lead you through a step-by-step guide of how to integrate your product to processing natural language using the API.AI tools. [block:api-header] { "type": "basic", "title": "Step 1: Create agent" } [/block] An API.AI agent represents a conversational interface of your application, device, or bot. You'll be prompted to create an agent right after you register at API.AI. If you are registered already, just <a href="https://console.api.ai/api-client/#/newAgent" target="_blank">follow this link</a> to create a new agent. [block:image] { "images": [ { "image": [ "https://files.readme.io/7a68a45-getting-started-create-agent.png", "getting-started-create-agent.png", 2134, 654, "#3482ac" ], "sizing": "full" } ] } [/block] After entering the name of your new agent, choose the language you want your app to communicate in. Currently, API.AI supports <a href="https://docs.api.ai/docs/languages" target="_blank">15 languages</a>. [block:callout] { "type": "warning", "body": "The language you choose for your agent cannot be changed after saving. However, you can always create a new agent if you wish to work in an additional language." } [/block] [block:image] { "images": [ { "image": [ "https://files.readme.io/c47e862-getting-started-select-lang.png", "getting-started-select-lang.png", 2090, 710, "#fafafa" ], "sizing": "full" } ] } [/block] Once you have named your agent and selected a language, click the 'Save' button. For more detailed information about agents, read <a href="https://docs.api.ai/docs/concept-agents" target="_blank">this documentation</a>. [block:api-header] { "type": "basic", "title": "Step 2: Create entities" } [/block] Once you’ve created your agent, you may want to create a few entities to extract necessary information from user queries. These will be stored in the <a href="https://console.api.ai/api-client/#/entities" target="_blank">Entities panel</a> located on the left of the API.AI developer console. [block:image] { "images": [ { "image": [ "https://files.readme.io/6932fb3-getting-started-entities.png", "getting-started-entities.png", 2058, 714, "#2a8ab9" ], "sizing": "full" } ] } [/block] Entities represent concepts or objects that are often specific to a use case. They are used for storing possible parameter values (along with their synonyms) that need to be returned in the <a href="https://docs.api.ai/docs/query#response" target="_blank">JSON response to a query</a>. The following example from a pizza delivery agent shows the entity titled “type”. It represents different types of pizza that the user can order. One pizza type can have several name variations. All such variations (synonyms) are mapped to a reference value that will be returned in the JSON response (unless other parameter value type is defined). [block:image] { "images": [ { "image": [ "https://files.readme.io/7d35454-getting-started-entity-example.png", "getting-started-entity-example.png", 2048, 1060, "#2e8ab3" ], "sizing": "full" } ] } [/block] Read more about <a href="https://docs.api.ai/docs/concept-entities" target="_blank">entities</a> and <a href="https://docs.api.ai/docs/concept-actions" target="_blank">parameters</a>. [block:api-header] { "type": "basic", "title": "Step 3: Create intents" } [/block] Once you created one or more entities, proceed to the <a href="https://console.api.ai/api-client/#/intents" target="_blank">Intents panel</a>. An intent determines what kind of user requests should be understood and to what actionable data (<a href="https://docs.api.ai/docs/query#response" target="_blank">JSON response</a>) those requests should be converted. Below is an example intent from a pizza delivery app, which is designed to understand some basic requests from users when they want to start their order. [block:image] { "images": [ { "image": [ "https://files.readme.io/fafb609-Getting-started_intent.png", "Getting-started_intent.png", 2044, 1044, "#f6f5f7" ], "sizing": "full" } ] } [/block] Note that the <a href="https://docs.api.ai/docs/machine-learning-settings" target="_blank">Machine Learning</a> option is enabled in all intents by default, which means that your agent will use request examples from the intent to train itself to understand many other variations of similar inputs. The more of the relevant and diverse examples you add to the intents, the smarter your agent will be. Read more about intents <a href="https://docs.api.ai/docs/concept-intents" target="_blank">here</a>. [block:api-header] { "type": "basic", "title": "Step 4: Test and train your agent" } [/block] In the top right corner of the API.AI developer console, you can find the test console, where you can test how well your agent understands you. [block:image] { "images": [ { "image": [ "https://files.readme.io/6146708-getting-started-test.png", "getting-started-test.png", 2790, 714, "#2894ad" ], "sizing": "full" } ] } [/block] If your request is recognized, you'll see which intent was used to process it and what information was extracted. [block:image] { "images": [ { "image": [ "https://files.readme.io/0e91ce6-getting-started-testing-results.png", "getting-started-testing-results.png", 554, 1248, "#f4f5f5" ] } ] } [/block] If your request was not matched to a desired intent or was matched incorrectly, you can add it to one of your intents right away or use the Training tool later to find all the unmatched inputs and use them to train your agent. You can find more information about Training <a href="https://docs.api.ai/docs/training" target="_blank">here</a>. [block:api-header] { "type": "basic", "title": "Step 5: Integrate" } [/block] Once your agent is ready, integrate it into your app or device. To facilitate this process, we created a number of useful tools: * <a href="https://docs.api.ai/docs/sdks" target="_blank">SDK's</a> for the most popular languages and platforms * <a href="https://docs.api.ai/docs/integrations" target="_blank">One-click integrations</a> for the most popular messaging platforms * <a href="https://docs.api.ai/docs/webhook" target="_blank">Webhook</a> integration with your web service