대화형 에이전트(Dialogflow CX)는 최종 사용자와의 동시 실행 대화를 처리하는 가상 에이전트입니다.
이는 인간 언어의 미묘한 차이를 이해하는 자연어 이해 모듈입니다.
대화형 에이전트(Dialogflow CX)는 대화 도중 앱과 서비스에서 최종 사용자의 텍스트나 오디오를 이해할 수 있는 정형 데이터로 변환합니다.
시스템에 필요한 대화 유형을 처리하도록 대화형 에이전트(Dialogflow CX) 에이전트를 직접 설계하고 빌드할 수 있습니다.
대화형 에이전트(Dialogflow CX) 에이전트는 콜센터 상담사와 비슷합니다.
둘 다 예상되는 대화 시나리오를 처리하도록 학습해야 하며, 학습이 지나치게 명시적일 필요는 없습니다.
에이전트 복원 시 다음을 제외한 모든 대상 에이전트 데이터(모든 흐름 버전 포함)를 덮어씁니다.
환경: 모든 커스텀 환경은 대상 에이전트에서 변경되지 않습니다.
대상 에이전트의 커스텀 환경에서 참조하는 흐름 버전은 연결된 환경이 존재하는 한 계속 존재합니다.
그러나 이러한 비활성 흐름 버전은 나열되지 않거나 에이전트에 대해 선택 가능한 흐름 버전이 아닙니다.
Vertex AI Agents 앱: Vertex AI Agents 앱과의 연결은 대상 에이전트에서 변경되지 않습니다. (즉 GenAppBuilderSettings의 engine 값) 결과 에이전트는 Vertex AI Agents 앱에도 연결해야 하므로 이는 데이터 스토어 에이전트는 다른 기존 데이터 스토어 에이전트에만 복원할 수 있다는 것을 의미합니다.
대상 에이전트가 앱과 연결되지 않은 경우 데이터 스토어 참조로 에이전트를 복원할 수 없습니다. 이렇게 하면 오류 메시지가 표시됩니다. 이 문제를 해결하려면 처음부터 새 데이터 스토어 에이전트를 만드세요. (또는 데이터 스토어 상태 핸들러를 추가하여 기존 에이전트를 데이터 스토어 에이전트로 전환할 수 있습니다. 이 경우 에이전트에서 관련 앱을 추가하는 방법이 안내됩니다.)
대상 에이전트가 앱과 연결된 경우 복원 시 모든 데이터 스토어 참조가 업데이트됩니다. Google Cloud 프로젝트 ID와 위치는 대상 에이전트의 앱과 일치하도록 업데이트됩니다. 컬렉션 ID와 데이터 스토어 ID는 변경되지 않습니다. 즉, 복원 작업 전에 대상 에이전트의 앱에 일치하는 유형의 모든 ID에 대한 데이터 스토어를 추가해야 합니다.
예를 들어 소스 에이전트가 projects/123/locations/eu-west2/collections/default_collection/dataStores/myDataStore1이라는 데이터 스토어를 참조하고 대상 에이전트의 앱 이름이 projects/321/locations/us-east1/collections/default_collections/engines/app123인 경우 결과 대상 에이전트의 데이터 스토어 참조는 projects/321/locations/us-east1/collections/default_collection/dataStores/myDataStore1이 됩니다.
내보낼 때 내보내기 파일 형식을 선택할 수 있습니다.
에이전트 데이터에 소스 제어 버전 관리를 사용하는 경우에는 JSON 형식으로 내보내야 합니다.
에이전트를 복원하면 대화형 에이전트(Dialogflow CX)에서 파일 형식을 자동으로 결정합니다.
[[["이해하기 쉬움","easyToUnderstand","thumb-up"],["문제가 해결됨","solvedMyProblem","thumb-up"],["기타","otherUp","thumb-up"]],[["이해하기 어려움","hardToUnderstand","thumb-down"],["잘못된 정보 또는 샘플 코드","incorrectInformationOrSampleCode","thumb-down"],["필요한 정보/샘플이 없음","missingTheInformationSamplesINeed","thumb-down"],["번역 문제","translationIssue","thumb-down"],["기타","otherDown","thumb-down"]],["최종 업데이트: 2025-09-04(UTC)"],[[["\u003cp\u003eConversational Agents (Dialogflow CX) agents are virtual agents that utilize natural language understanding to handle concurrent conversations with end-users, translating text or audio into structured data.\u003c/p\u003e\n"],["\u003cp\u003eCreating an agent can be done via the Dialogflow CX console or API, with options to auto-generate a data store agent or build a custom agent.\u003c/p\u003e\n"],["\u003cp\u003eAgents serve as containers for virtual agent data like intents, entity types, webhooks, flows, pages, and route groups.\u003c/p\u003e\n"],["\u003cp\u003eAgents can be exported to a file (excluding flow versions and custom environments) and restored, overwriting existing data, with specific handling for data store agent associations and Vertex AI Agents Apps.\u003c/p\u003e\n"],["\u003cp\u003eDeleting an agent is permanent, requiring proper permissions and a backup via export is recommended, and deleting a project will immediately delete all associated agents.\u003c/p\u003e\n"]]],[],null,["# Agents\n\nA\n\n*Conversational Agents (Dialogflow CX) agent*\n\nis a virtual agent\nthat handles concurrent conversations with your end-users.\nIt is a natural language understanding module\nthat understands the nuances of human language.\nConversational Agents (Dialogflow CX) translates end-user text or audio during a conversation\nto structured data that your apps and services can understand.\nYou design and build a Conversational Agents (Dialogflow CX) agent\nto handle the types of conversations required for your system.\n\nA Conversational Agents (Dialogflow CX) agent is similar to a human call center agent.\nYou train them both to handle expected conversation scenarios,\nand your training does not need to be overly explicit.\n\nCreate an agent\n---------------\n\n| **Note:** You can create multiple Conversational Agents (Dialogflow CX) agents for one [Google Cloud project](/resource-manager/docs/creating-managing-projects).\n\nTo create an agent: \n\n### Console\n\n1. Open the [Dialogflow CX console](https://dialogflow.cloud.google.com/cx/projects).\n2. Create or choose a Google Cloud project.\n3. Click **Create agent**.\n4. Select **Auto-generate** to create a [data store agent](/dialogflow/cx/docs/concept/data-store-agent) or select **Build your own** to create other kinds of agents.\n5. Complete the form for basic agent settings:\n 1. You can choose any display name.\n 2. Select your preferred [location](/dialogflow/cx/docs/how/region#avail). Click the **Edit** button if you want to change advanced [location settings](/dialogflow/cx/docs/how/region#location-settings).\n 3. Select your preferred time zone.\n 4. Select the default language for your agent. You cannot change the default language for an agent once it is created.\n6. Click **Save**.\n\n### API\n\nIf you have not already configured\n[location settings](/dialogflow/cx/docs/concept/region#location-settings)\nfor your project,\nyou must configure these settings with the console\nbefore creating agents with the API.\nCurrently, you cannot configure location settings with the API.\n\nTo create an agent,\nsee the `create` method for the `Agent` type.\n\n\nGo to the Agent API reference \n**Select a protocol and version for the Agent reference:**\n\nClose\n\n\u003cbr /\u003e\n\nAgent data\n----------\n\nConversational Agents (Dialogflow CX) agents serve as top-level containers\nfor settings and data for virtual agents.\nThe following data is associated with agents:\n\n- [Intents](/dialogflow/cx/docs/concept/intent)\n- [Entity types](/dialogflow/cx/docs/concept/entity)\n- [Webhooks](/dialogflow/cx/docs/concept/webhook)\n- [Flows](/dialogflow/cx/docs/concept/flow)\n- [Pages](/dialogflow/cx/docs/concept/page)\n- [Route groups](/dialogflow/cx/docs/concept/handler#group)\n\nFor more information about how data is applied at varying levels, see the\n[data application levels](/dialogflow/cx/docs/concept/data-level).\n\nExport and restore an agent\n---------------------------\n\n| **Warning:** We will no longer export raw value credentials for OpenAPI Tools and Webhooks, starting Aug 15, 2025. You should migrate to store your credentials in Secret Manager. See [Webhook](/dialogflow/cx/docs/concept/webhook#secret-manager-auth) and [Tool](/dialogflow/cx/docs/concept/playbook/tool#secret-manager-auth) documentations for instructions.\n\nYou can export an agent to a file,\nand restore an agent with that file.\n\nAn agent export includes all agent data except the following:\n\n- [Flow versions](/dialogflow/cx/docs/concept/version): Only the draft flows are exported to file.\n- [Environments](/dialogflow/cx/docs/concept/version): Custom environments are not exported to file.\n\nAn agent restore overwrites all target agent data\n(including all flow versions) except the following:\n\n- [Environments](/dialogflow/cx/docs/concept/version): All custom environments remain unchanged in the target agent. Flow versions referenced by custom environments in the target agent will continue to exist, as long as the associated environments exist. However, these stale flow versions are not listed or selectable flow versions for the agent.\n- [Vertex AI Agents Apps](/generative-ai-app-builder/docs/agent-intro): The association to a Vertex AI Agents App remains unchanged in the target agent. (In other words, the value of `engine` in [GenAppBuilderSettings](/dialogflow/cx/docs/reference/rest/v3/projects.locations.agents#GenAppBuilderSettings)) This means that data store agents can only be restored into other existing data store agents, because the resulting agent also needs to have an association to a Vertex AI Agents App.\n- [Vertex AI Agents Data Stores](/generative-ai-app-builder/docs/agent-usage):\n All references to data stores will be overwritten in the target agent\n according to the following rules:\n\n - If the target agent isn't associated with an App, then it's not possible to restore an agent with data store references into it. Trying to do so results in an error message. To fix that, you can either [create a new data store agent](/generative-ai-app-builder/docs/agent-usage#create_a_data_store_agent) from scratch. (Alternatively, you can turn your existing agent into a data store agent by adding a data store [state handler](/dialogflow/cx/docs/concept/handler) to it. In this case you'll be guided through adding an associated App to your agent.)\n - If the target agent is associated with an App, then all the data store references will be updated upon restore: their Google Cloud project ID and location will be updated to match the App of the target agent. The collection ID and data store ID will remain unchanged. This means that you need to add data stores for all the IDs with matching types into the App of the target agent prior to the restore operation.\n\n Example: if the source agent refers to a data store named\n `projects/123/locations/eu-west2/collections/default_collection/dataStores/myDataStore1`\n and the App of the target agent is named\n `projects/321/locations/us-east1/collections/default_collections/engines/app123`,\n then the resulting data store reference in the target agent will become:\n `projects/321/locations/us-east1/collections/default_collection/dataStores/myDataStore1`\n\n| **Note:** in the API and in the contents of the exported data, a Vertex AI Agents App is called a GenAppBuilder Engine. For example in an exported JSON Package you can find the name of the engine in the `agent.json` file under the key `genAppBuilderSettings.engine`.\n\nWhen exporting,\nyou can select the export file format.\nIf you are using source control versioning for your agent data,\nyou should\n[export in the JSON format](/dialogflow/cx/docs/reference/json-export).\nWhen you restore an agent,\nConversational Agents (Dialogflow CX) automatically determines the file format.\n\nTo export or restore an agent: \n\n### Console\n\n1. Open the [Dialogflow CX console](https://dialogflow.cloud.google.com/cx/projects).\n2. Choose the Google Cloud project for the agent.\n3. Click the option *more_vert* menu for an agent in the list.\n4. Click the **Export** or **Restore** button.\n5. Follow instructions to complete.\n**Note:** To restore a data store agent, make sure that the target agent was created as a [data store agent](/generative-ai-app-builder/docs/agent-usage#create_a_data_store_agent). \n\n### API\n\nSee the `export` and `restore` methods for the `Agent` type.\n\n\nGo to the Agent API reference \n**Select a protocol and version for the Agent reference:**\n\nClose\n\n\u003cbr /\u003e\n\nIf the agent size exceeds the [maximum limit](/dialogflow/quotas#size), use the\nCloud Storage option for agent export and restore.\n\nIf you use GitHub, also see the\n[GitHub export/restore guide](/dialogflow/cx/docs/concept/github).\n\nDelete an agent\n---------------\n\n| **Caution:** Deleting an agent **cannot** be undone. [Export](#export) your agent to keep a backup if necessary.\n\nIn order to delete an agent,\nyou need a role that provides full access or edit access.\nSee the\n[access control guide](/dialogflow/cx/docs/concept/access-control)\nfor more information.\n\nTo delete an agent: \n\n### Console\n\n1. Open the [Dialogflow CX console](https://dialogflow.cloud.google.com/cx/projects).\n2. Choose the Google Cloud project for the agent.\n3. Click the option *more_vert* menu for an agent in the list.\n4. Click the delete *delete* button.\n5. Confirm deletion in the dialog.\n\n### API\n\nSee the `delete` method for the `Agent` type.\n\n\nGo to the Agent API reference \n**Select a protocol and version for the Agent reference:**\n\nClose\n\n\u003cbr /\u003e\n\nIf you\n[delete your project](/resource-manager/docs/creating-managing-projects#shutting_down_projects),\nall Conversational Agents (Dialogflow CX) agents and data associated with the project\nare deleted immediately."]]
