{"allowedRenditionsWidth":["320","480","640","768","960","1200","1440","1920"],"templateName":"quickstart-page-template","cssClassNames":"page basicpage summit-page","canonicalLink":"https://www.snowflake.com/en/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/","robotsTags":[],"language":"en","title":"Build Production AI via Natural Language with Dataiku Cobuild on Snowflake","analyticsPageType":"quickstart-page-template","analyticsCategory":"general","analyticsSubCategory":"","excludeFromAnalytics":false,"coveoConfig":{"organizationId":"snowflakecomputingproduction8neljofn","apiKey":"xx335921a6-2a0a-40f2-a167-e390b4766c3d","pipeline":"snowflake.com","searchHub":"snowflake.com"},"analyticsDebugMode":false,"analyticsData":{"excludeFromAnalytics":false,"subCategory":"","pageType":"quickstart-page-template","templateName":"quickstart-page-template","siteName":"snowflake","pageUrl":"/content/snowflake-site/global/en/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake","language":"en","category":"general","pageName":"Build Production AI via Natural Language with Dataiku Cobuild on Snowflake","contentTags":["snowflake-site:taxonomy/solution-center/certification/quickstart","snowflake-site:taxonomy/product/ai","snowflake-site:taxonomy/snowflake-feature/model-development","snowflake-site:taxonomy/solution-center/certification/partner-solution","snowflake-site:taxonomy/snowflake-feature/applied-analytics"]},":items":{"root":{"columnClassNames":{"markup_editor_1950346551":"aem-GridColumn aem-GridColumn--default--12","experiencefragment-banner":"aem-GridColumn aem-GridColumn--default--12","experiencefragment-header":"aem-GridColumn aem-GridColumn--default--12","responsivegrid":"aem-GridColumn aem-GridColumn--default--12","experiencefragment-footer":"aem-GridColumn aem-GridColumn--default--12","modal_container":"aem-GridColumn aem-GridColumn--default--12","markup_editor":"aem-GridColumn aem-GridColumn--default--12"},"gridClassNames":"aem-Grid aem-Grid--12 aem-Grid--default--12","columnCount":12,":items":{"experiencefragment-banner":{"id":"experiencefragment-d033c3fbb3","localizedFragmentVariationPath":"/content/experience-fragments/snowflake-site/language-masters/en/site/pushdown-banner/master/jcr:content","configured":true,"xfModelPath":"/content/experience-fragments/snowflake-site/language-masters/en/site/pushdown-banner/master.xfmodel.json",":type":"snowflake-site/components/experiencefragment"},"experiencefragment-header":{"id":"experiencefragment-7b4d0aeaad","localizedFragmentVariationPath":"/content/experience-fragments/snowflake-site/language-masters/en/site/mega-nav-header/master/jcr:content","configured":true,"xfModelPath":"/content/experience-fragments/snowflake-site/language-masters/en/site/mega-nav-header/master.xfmodel.json","languageNavPath":"/content/snowflake-site/global/en/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake.languagenav.json",":type":"snowflake-site/components/experiencefragment"},"markup_editor_1950346551":{"id":"markup-editor-a16f26a333","title":" ","cssContent":".snowflake-markdown-table code[class*=language-],.snowflake-markdown-table code[class*=language-],.snowflake-markdown .snowflake-text code[class*=language-],.snowflake-markdown .snowflake-text pre[class*=language-]{background-color:rgba(var(--ui-12-rgb),.5);color:var(--text-01);text-shadow:none;padding:var(--spacing-00);border-radius:var(--spacing-00);font-size:smaller}","isGSAPEnabled":false,":type":"snowflake-site/components/markup-editor"},"responsivegrid":{"columnClassNames":{"quickstart_hero":"aem-GridColumn aem-GridColumn--default--12","flexible_column_cont":"aem-GridColumn aem-GridColumn--default--12","markup_editor":"aem-GridColumn aem-GridColumn--default--12"},"gridClassNames":"aem-Grid aem-Grid--12 aem-Grid--default--12","columnCount":12,":items":{"quickstart_hero":{"id":"quickstart-hero-67dad3eb0d","fragmentPath":"/content/dam/snowflake-site/en/content-fragments/quickstarts/build-production-ai-with-dataiku-cobuild-on-snowflake","isDeveloperGuidesPage":false,"quickstartHeroFirstCertifiedTag":{"tagText":"Partner Solution","tagColor":"#29B5E8","tagPath":"/content/cq:tags/snowflake-site/taxonomy/solution-center/certification/partner-solution","tagIcon":""},"quickstartHeroTitle":{"lines":["Build Production AI via Natural Language with Dataiku Cobuild on Snowflake"],"type":"heading2",":type":"snowflake-site/components/title-v2"},"quickstartHeroAuthor":"Randy Pettus","quickstartHeroFirstSnowflakeFeatureTag":{"tagText":"Applied Analytics","tagColor":"#29B5E8","tagPath":"/content/cq:tags/snowflake-site/taxonomy/snowflake-feature/applied-analytics","tagIcon":""},"quickstartHeroForkRepoLink":{"id":"button-e5b435a893","showOutboundIcon":false,"buttonLink":{"valid":true,"attributes":{"target":"_blank"},"url":"https://github.com/Snowflake-Labs/sfquickstarts/tree/master/site/sfguides/src/build-production-ai-with-dataiku-cobuild-on-snowflake"},"linkTargetContentType":"GENERIC","linkType":"SNOWFLAKE_EXTERNAL",":type":"snowflake-site/components/button","text":"Fork Repo"},":type":"snowflake-site/components/quickstart/quickstart-hero","quickstartHeroBreadcrumbs":[{"title":"Build Production AI via Natural Language with Dataiku Cobuild on Snowflake","url":"https://www.snowflake.com/content/snowflake-site/global/en/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake","currentPage":true},{"title":"Guides","url":"https://www.snowflake.com/content/snowflake-site/global/en/developers/guides","currentPage":false},{"title":"Snowflake for Developers","url":"https://www.snowflake.com/content/snowflake-site/global/en/developers","currentPage":false}]},"flexible_column_cont":{"id":"flexible-column-container-8a326e7a53","propertiesId":"quickstart-template-main-flexible-container","type":"2-column-75-25","alignColumns":"top","containerMaxWidth":"extra-large","topPadding":"none","bottomPadding":"none","spaceBetween":"small","reverseOnMobile":false,"carouselOnMobile":false,"backgroundImageOption":"none","flexible_column_content_container_1":{"layout":"SIMPLE","id":"container-aa4518e768",":items":{"contentfragment":{"id":"contentfragment-cdad7001d6","description":"","paragraphs":["<!-- ------------------------ -->\n","\u003Ch2\u003EOverview\u003C/h2\u003E\n","\u003Cp\u003EDuration: 3\u003C/p\u003E\n","\u003Cp\u003E\u003Ca href=\"https://www.dataiku.com/product/cobuild-on-snowflake\"\u003EDataiku Cobuild on Snowflake\u003C/a\u003E is an AI agent that turns natural-language requests into production-ready AI workflows, agents, and applications running on Snowflake. Powered by Snowflake Cortex, Cobuild lets teams describe a goal in plain language and get back an inspectable visual workflow they can review and govern before it reaches production, putting AI building within reach of analysts, operations teams, and business users.\u003C/p\u003E\n","\u003Cp\u003EThis guide walks through how to set up Dataiku Cobuild with Snowflake Cortex to give your team model flexibility and AI governance while building production-grade data and ML pipelines. Cobuild is particularly useful for:\u003C/p\u003E\n\u003Cul\u003E\u003Cli\u003E\u003Cstrong\u003EBuilding new data and AI projects from scratch:\u003C/strong\u003E describe your goal in plain language and Cobuild generates a complete, visual flow ready for review and deployment\u003C/li\u003E\u003Cli\u003E\u003Cstrong\u003ERefactoring existing workflows:\u003C/strong\u003E point Cobuild at an existing project and ask it to optimize, restructure, or extend what's already there\u003C/li\u003E\u003Cli\u003E\u003Cstrong\u003EWriting documentation:\u003C/strong\u003E ask Cobuild to explain any project or pipeline in plain language and write detailed documentation, making it easier for teams to understand and hand off work\u003C/li\u003E\u003C/ul\u003E\n","\u003Cp\u003E\u003Cimg src=\"https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/cobuild_architecture.png\" alt=\"Cobuild on Cortex Architecture\"\u003E\u003C/p\u003E\n","\u003Ch3\u003EPrerequisites\u003C/h3\u003E\n","\u003Cp\u003EThis guide is designed for users who already have a Dataiku instance connected to Snowflake. If you are new to Dataiku and Snowflake, visit the \u003Ca href=\"https://www.snowflake.com/en/developers/guides/\"\u003ESnowflake Developers Guides\u003C/a\u003E and search for Dataiku to find guides that cover the initial setup and integration.\u003C/p\u003E\n\u003Cul\u003E\u003Cli\u003EA Snowflake account with access to Cortex LLM features\u003C/li\u003E\u003Cli\u003EA Dataiku instance connected to Snowflake\u003C/li\u003E\u003C/ul\u003E\n","\u003Ch3\u003EWhat You'll Learn\u003C/h3\u003E\n\u003Cul\u003E\u003Cli\u003EHow to prepare your Snowflake environment for Cobuild\u003C/li\u003E\u003Cli\u003EHow to connect Snowflake Cortex to Dataiku\u003C/li\u003E\u003Cli\u003EHow to use Dataiku Cobuild to build and deploy production AI pipelines via natural language\u003C/li\u003E\u003C/ul\u003E\n","\u003Ch3\u003EWhat You'll Need\u003C/h3\u003E\n\u003Cul\u003E\u003Cli\u003EACCOUNTADMIN or equivalent privileges in Snowflake to configure role grants and account-level settings\u003C/li\u003E\u003Cli\u003EAn active Dataiku instance with an existing Snowflake connection and administrative access\u003C/li\u003E\u003C/ul\u003E\n","\u003Ch3\u003EWhat You'll Build\u003C/h3\u003E\n\u003Cul\u003E\u003Cli\u003EA production AI pipeline built through natural language prompts in Dataiku Cobuild, governed and deployed on Snowflake\u003C/li\u003E\u003C/ul\u003E\n<!-- ------------------------ -->\n","\u003Ch2\u003ESnowflake Setup Steps\u003C/h2\u003E\n","\u003Cp\u003EDuration: 2\u003C/p\u003E\n","\u003Ch3\u003ECortex Role Grants\u003C/h3\u003E\n","\u003Cp\u003EDataiku calls Snowflake Cortex via the REST API, which requires the user's default role to have either \u003Ccode\u003ESNOWFLAKE.CORTEX_USER\u003C/code\u003E (all Cortex features, granted to PUBLIC by default) or \u003Ccode\u003ESNOWFLAKE.CORTEX_REST_API_USER\u003C/code\u003E (REST API only). In most accounts no action is needed. If access has been revoked, run as ACCOUNTADMIN:\u003C/p\u003E\n\u003Cpre\u003E\u003Ccode class=\"language-sql\"\u003E-- Option 1: Full Cortex access\nGRANT DATABASE ROLE SNOWFLAKE.CORTEX_USER TO ROLE <your_dataiku_role>;\n\n-- Option 2: REST API access only\nGRANT DATABASE ROLE SNOWFLAKE.CORTEX_REST_API_USER TO ROLE <your_dataiku_role>;\n\u003C/code\u003E\u003C/pre\u003E\n","\u003Cp\u003EAlso confirm the Dataiku service user's default role is set correctly:\u003C/p\u003E\n\u003Cpre\u003E\u003Ccode class=\"language-sql\"\u003EALTER USER <dataiku_service_user> SET DEFAULT_ROLE = <your_dataiku_role>;\n\u003C/code\u003E\u003C/pre\u003E\n","\u003Cp\u003ESee the \u003Ca href=\"https://docs.snowflake.com/en/user-guide/snowflake-cortex/cortex-rest-api\"\u003ESnowflake Cortex REST API documentation\u003C/a\u003E for more detail.\u003C/p\u003E\n","\u003Ch3\u003EEnable Cortex Cross-Region Inference\u003C/h3\u003E\n","\u003Cp\u003EEnabling cross-region inference gives Dataiku access to a broader set of foundation models and increases rate limits by routing requests to available inference capacity across regions.\u003C/p\u003E\n","\u003Cp\u003ERun the following as ACCOUNTADMIN:\u003C/p\u003E\n\u003Cpre\u003E\u003Ccode class=\"language-sql\"\u003EALTER ACCOUNT SET CORTEX_ENABLED_CROSS_REGION = 'ANY_REGION';\n\u003C/code\u003E\u003C/pre\u003E\n<!-- ------------------------ -->\n","\u003Ch2\u003EConnecting Cortex to Dataiku\u003C/h2\u003E\n","\u003Cp\u003EDuration: 5\u003C/p\u003E\n","\u003Ch3\u003ECreate a New Snowflake Cortex Connection\u003C/h3\u003E\n","\u003Cp\u003EIn Dataiku, navigate to \u003Cstrong\u003EAdministration > Connections\u003C/strong\u003E, click \u003Cstrong\u003E+ New Connection\u003C/strong\u003E, and select \u003Cstrong\u003ESnowflake Cortex\u003C/strong\u003E from the connection type list.\u003C/p\u003E\n","\u003Cp\u003E\u003Cimg src=\"https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/cortex_connection.png\" alt=\"Select Snowflake Cortex\"\u003E\u003C/p\u003E\n","\u003Ch3\u003EConfigure the Connection\u003C/h3\u003E\n","\u003Cp\u003ESelect your existing Snowflake connection and configure your model options. From this screen you can control which models are available to your team. Under \u003Cstrong\u003EUsage Control\u003C/strong\u003E, you can enable \u003Cstrong\u003EPrompt Caching\u003C/strong\u003E to reduce redundant token usage and lower inference costs.\u003C/p\u003E\n","\u003Cp\u003E\u003Cimg src=\"https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/cortex_config.png\" alt=\"Cortex Connection Config\"\u003E\u003C/p\u003E\n","\u003Cp\u003EOnce saved, this connection is available across LLM Mesh, Agents, LLM recipes, and Cobuild.\u003C/p\u003E\n","\u003Ch3\u003EEnable AI Services for Cobuild\u003C/h3\u003E\n","\u003Cp\u003EWith the Cortex connection in place, you need to point Dataiku's AI Services at it so Cobuild can use it. From the grid menu in the top navigation, go to \u003Cstrong\u003EAdministration\u003C/strong\u003E.\u003C/p\u003E\n","\u003Cp\u003E\u003Cimg src=\"https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/cobuild_admin.png\" alt=\"Dataiku Administration Menu\"\u003E\u003C/p\u003E\n","\u003Cp\u003EClick the \u003Cstrong\u003EAI Services\u003C/strong\u003E tab, set the Operation mode to \u003Cstrong\u003EBring your own LLM\u003C/strong\u003E, and click \u003Cstrong\u003EEnable\u003C/strong\u003E. You will be prompted to accept Dataiku's AI Services Terms of Use before proceeding.\u003C/p\u003E\n","\u003Cp\u003E\u003Cimg src=\"https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/cobuild_ai_services.png\" alt=\"AI Services Settings\"\u003E\u003C/p\u003E\n","\u003Ch3\u003EConfigure Cobuild LLMs\u003C/h3\u003E\n","\u003Cp\u003EUnder \u003Cstrong\u003ELLMs to use\u003C/strong\u003E, enter your \u003Cstrong\u003EReference project key\u003C/strong\u003E (this is used for LLM Mesh requests that run outside a DSS project, such as AI Search). Then select a \u003Cstrong\u003EMain model\u003C/strong\u003E from your Snowflake Cortex connection. You can also optionally assign a \u003Cstrong\u003EFast / light model\u003C/strong\u003E and a \u003Cstrong\u003EResponse-format-aware model\u003C/strong\u003E for different task types.\u003C/p\u003E\n","\u003Cp\u003E\u003Cimg src=\"https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/configure_cobuild.png\" alt=\"Configure Cobuild LLMs\"\u003E\u003C/p\u003E\n","\u003Cp\u003EFill in any additional usage permissions or global custom instructions, then click \u003Cstrong\u003ESave\u003C/strong\u003E. Your Cortex configuration is now active and ready for Cobuild.\u003C/p\u003E\n","\u003Cp\u003E\u003Cimg src=\"https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/cobuild_save_ai_services.png\" alt=\"Save AI Services\"\u003E\u003C/p\u003E\n<!-- ------------------------ -->\n","\u003Ch2\u003EUsing Dataiku Cobuild\u003C/h2\u003E\n","\u003Cp\u003EDuration: 5\u003C/p\u003E\n","\u003Ch3\u003EStart with an Existing Project\u003C/h3\u003E\n","\u003Cp\u003EOpen any existing Dataiku project and click \u003Cstrong\u003ECobuild\u003C/strong\u003E from the project view.\u003C/p\u003E\n","\u003Cp\u003E\u003Cimg src=\"https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/cobuild_project_view_selection.png\" alt=\"Cobuild Project View\"\u003E\u003C/p\u003E\n","\u003Cp\u003ECobuild will inspect your flow automatically. You can also select specific items directly in the flow to give Cobuild more targeted context, or type \u003Ccode\u003E@\u003C/code\u003E in the prompt to reference a specific dataset, recipe, or model by name.\u003C/p\u003E\n","\u003Cp\u003EBefore running any actions, you can choose how Cobuild executes: either prompt you to \u003Cstrong\u003Eapprove each action\u003C/strong\u003E before it runs, or set it to \u003Cstrong\u003Eautomatic execution\u003C/strong\u003E so Cobuild applies changes without waiting for confirmation.\u003C/p\u003E\n","\u003Cp\u003ETo get started with Cobuild, type:\u003C/p\u003E\n\u003Cpre\u003E\u003Ccode\u003EExplain my project\n\u003C/code\u003E\u003C/pre\u003E\n","\u003Cp\u003ECobuild will analyze your project and return a plain-language summary of what it contains, how it's structured, and what it does. Here's an example of what that looks like:\u003C/p\u003E\n","\u003Cp\u003E\u003Cimg src=\"https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/cobuild_explain.png\" alt=\"Cobuild Project Explanation\"\u003E\u003C/p\u003E\n","\u003Cp\u003EThis is a good starting point to verify your Cortex connection is working and to get oriented before building or refactoring.\u003C/p\u003E\n","\u003Ch3\u003EBuild a Pipeline from Scratch\u003C/h3\u003E\n","\u003Cp\u003EOnce you're comfortable with how Cobuild reads your project, try starting fresh with a new dataset connected to Snowflake. Connect your data and then ask Cobuild to do the heavy lifting:\u003C/p\u003E\n\u003Cpre\u003E\u003Ccode\u003EBuild a full end-to-end pipeline using this data. Prepare the features, train a model, and score it on unseen data.\n\u003C/code\u003E\u003C/pre\u003E\n","\u003Cp\u003ECobuild will generate the complete flow as a visual, inspectable pipeline that your team can review and refine before anything runs in production.\u003C/p\u003E\n<!-- ------------------------ -->\n","\u003Ch2\u003EConclusion And Resources\u003C/h2\u003E\n","\u003Cp\u003EDuration: 2\u003C/p\u003E\n","\u003Cp\u003EYou now have Dataiku Cobuild connected to Snowflake Cortex and a production AI pipeline built entirely through natural language. From here you can apply the same workflow to new projects, refactor existing Dataiku flows, or use Cobuild to generate documentation for pipelines your team already owns.\u003C/p\u003E\n","\u003Ch3\u003EWhat You Learned\u003C/h3\u003E\n\u003Cul\u003E\u003Cli\u003EHow to configure Snowflake for use with Dataiku Cobuild\u003C/li\u003E\u003Cli\u003EHow to connect Snowflake Cortex to a Dataiku instance\u003C/li\u003E\u003Cli\u003EHow to build and deploy a production AI pipeline using natural language in Cobuild\u003C/li\u003E\u003C/ul\u003E\n","\u003Ch3\u003ERelated Resources\u003C/h3\u003E\n\u003Cul\u003E\u003Cli\u003E\u003Ca href=\"https://www.dataiku.com/product/cobuild-on-snowflake\"\u003EDataiku Cobuild on Snowflake\u003C/a\u003E\u003C/li\u003E\u003Cli\u003E\u003Ca href=\"https://knowledge.dataiku.com/latest/ui/cobuild/concept-cobuild.html\"\u003EDataiku Cobuild Knowledge Base\u003C/a\u003E\u003C/li\u003E\u003Cli\u003E\u003Ca href=\"https://doc.dataiku.com\"\u003EDataiku Documentation\u003C/a\u003E\u003C/li\u003E\u003Cli\u003E\u003Ca href=\"https://docs.snowflake.com/en/user-guide/snowflake-cortex/cortex-rest-api\"\u003ESnowflake Cortex REST API\u003C/a\u003E\u003C/li\u003E\u003Cli\u003E\u003Ca href=\"https://docs.snowflake.com/en/guides-overview-ai-features\"\u003ESnowflake Cortex AI Documentation\u003C/a\u003E\u003C/li\u003E\u003Cli\u003E\u003Ca href=\"https://app.snowflake.com/marketplace/listing/GZTSZNUFX99I/dataiku-dataiku-for-enterprise-ai\"\u003EDataiku on Snowflake Marketplace\u003C/a\u003E\u003C/li\u003E\u003C/ul\u003E"],"title":"Base Quickstart CF",":items":{},":itemsOrder":[],"elements":{"quickstartArticleBody":{"dataType":"string","title":"Quickstart Article Body","value":"\u003C!-- ------------------------ --\u003E\n\n## Overview\nDuration: 3\n\n[Dataiku Cobuild on Snowflake](https://www.dataiku.com/product/cobuild-on-snowflake) is an AI agent that turns natural-language requests into production-ready AI workflows, agents, and applications running on Snowflake. Powered by Snowflake Cortex, Cobuild lets teams describe a goal in plain language and get back an inspectable visual workflow they can review and govern before it reaches production, putting AI building within reach of analysts, operations teams, and business users.\n\nThis guide walks through how to set up Dataiku Cobuild with Snowflake Cortex to give your team model flexibility and AI governance while building production-grade data and ML pipelines. Cobuild is particularly useful for:\n\n- **Building new data and AI projects from scratch:** describe your goal in plain language and Cobuild generates a complete, visual flow ready for review and deployment\n- **Refactoring existing workflows:** point Cobuild at an existing project and ask it to optimize, restructure, or extend what's already there\n- **Writing documentation:** ask Cobuild to explain any project or pipeline in plain language and write detailed documentation, making it easier for teams to understand and hand off work\n\n![Cobuild on Cortex Architecture](https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/cobuild_architecture.png)\n\n### Prerequisites\n\nThis guide is designed for users who already have a Dataiku instance connected to Snowflake. If you are new to Dataiku and Snowflake, visit the [Snowflake Developers Guides](https://www.snowflake.com/en/developers/guides/) and search for Dataiku to find guides that cover the initial setup and integration.\n\n- A Snowflake account with access to Cortex LLM features\n- A Dataiku instance connected to Snowflake\n\n### What You'll Learn\n\n- How to prepare your Snowflake environment for Cobuild\n- How to connect Snowflake Cortex to Dataiku\n- How to use Dataiku Cobuild to build and deploy production AI pipelines via natural language\n\n### What You'll Need\n\n- ACCOUNTADMIN or equivalent privileges in Snowflake to configure role grants and account-level settings\n- An active Dataiku instance with an existing Snowflake connection and administrative access\n\n### What You'll Build\n\n- A production AI pipeline built through natural language prompts in Dataiku Cobuild, governed and deployed on Snowflake\n\n\u003C!-- ------------------------ --\u003E\n\n## Snowflake Setup Steps\nDuration: 2\n\n### Cortex Role Grants\n\nDataiku calls Snowflake Cortex via the REST API, which requires the user's default role to have either `SNOWFLAKE.CORTEX_USER` (all Cortex features, granted to PUBLIC by default) or `SNOWFLAKE.CORTEX_REST_API_USER` (REST API only). In most accounts no action is needed. If access has been revoked, run as ACCOUNTADMIN:\n\n```sql\n-- Option 1: Full Cortex access\nGRANT DATABASE ROLE SNOWFLAKE.CORTEX_USER TO ROLE \u003Cyour_dataiku_role\u003E;\n\n-- Option 2: REST API access only\nGRANT DATABASE ROLE SNOWFLAKE.CORTEX_REST_API_USER TO ROLE \u003Cyour_dataiku_role\u003E;\n```\n\nAlso confirm the Dataiku service user's default role is set correctly:\n\n```sql\nALTER USER \u003Cdataiku_service_user\u003E SET DEFAULT_ROLE = \u003Cyour_dataiku_role\u003E;\n```\n\nSee the [Snowflake Cortex REST API documentation](https://docs.snowflake.com/en/user-guide/snowflake-cortex/cortex-rest-api) for more detail.\n\n### Enable Cortex Cross-Region Inference\n\nEnabling cross-region inference gives Dataiku access to a broader set of foundation models and increases rate limits by routing requests to available inference capacity across regions.\n\nRun the following as ACCOUNTADMIN:\n\n```sql\nALTER ACCOUNT SET CORTEX_ENABLED_CROSS_REGION = 'ANY_REGION';\n```\n\n\u003C!-- ------------------------ --\u003E\n\n## Connecting Cortex to Dataiku\nDuration: 5\n\n### Create a New Snowflake Cortex Connection\n\nIn Dataiku, navigate to **Administration \u003E Connections**, click **+ New Connection**, and select **Snowflake Cortex** from the connection type list.\n\n![Select Snowflake Cortex](https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/cortex_connection.png)\n\n### Configure the Connection\n\nSelect your existing Snowflake connection and configure your model options. From this screen you can control which models are available to your team. Under **Usage Control**, you can enable **Prompt Caching** to reduce redundant token usage and lower inference costs.\n\n![Cortex Connection Config](https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/cortex_config.png)\n\nOnce saved, this connection is available across LLM Mesh, Agents, LLM recipes, and Cobuild.\n\n### Enable AI Services for Cobuild\n\nWith the Cortex connection in place, you need to point Dataiku's AI Services at it so Cobuild can use it. From the grid menu in the top navigation, go to **Administration**.\n\n![Dataiku Administration Menu](https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/cobuild_admin.png)\n\nClick the **AI Services** tab, set the Operation mode to **Bring your own LLM**, and click **Enable**. You will be prompted to accept Dataiku's AI Services Terms of Use before proceeding.\n\n![AI Services Settings](https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/cobuild_ai_services.png)\n\n### Configure Cobuild LLMs\n\nUnder **LLMs to use**, enter your **Reference project key** (this is used for LLM Mesh requests that run outside a DSS project, such as AI Search). Then select a **Main model** from your Snowflake Cortex connection. You can also optionally assign a **Fast / light model** and a **Response-format-aware model** for different task types.\n\n![Configure Cobuild LLMs](https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/configure_cobuild.png)\n\nFill in any additional usage permissions or global custom instructions, then click **Save**. Your Cortex configuration is now active and ready for Cobuild.\n\n![Save AI Services](https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/cobuild_save_ai_services.png)\n\n\u003C!-- ------------------------ --\u003E\n\n## Using Dataiku Cobuild\nDuration: 5\n\n### Start with an Existing Project\n\nOpen any existing Dataiku project and click **Cobuild** from the project view.\n\n![Cobuild Project View](https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/cobuild_project_view_selection.png)\n\nCobuild will inspect your flow automatically. You can also select specific items directly in the flow to give Cobuild more targeted context, or type `@` in the prompt to reference a specific dataset, recipe, or model by name.\n\nBefore running any actions, you can choose how Cobuild executes: either prompt you to **approve each action** before it runs, or set it to **automatic execution** so Cobuild applies changes without waiting for confirmation.\n\nTo get started with Cobuild, type:\n\n```\nExplain my project\n```\n\nCobuild will analyze your project and return a plain-language summary of what it contains, how it's structured, and what it does. Here's an example of what that looks like:\n\n![Cobuild Project Explanation](https://www.snowflake.com/content/dam/snowflake-site/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/cobuild_explain.png)\n\nThis is a good starting point to verify your Cortex connection is working and to get oriented before building or refactoring.\n\n### Build a Pipeline from Scratch\n\nOnce you're comfortable with how Cobuild reads your project, try starting fresh with a new dataset connected to Snowflake. Connect your data and then ask Cobuild to do the heavy lifting:\n\n```\nBuild a full end-to-end pipeline using this data. Prepare the features, train a model, and score it on unseen data.\n```\n\nCobuild will generate the complete flow as a visual, inspectable pipeline that your team can review and refine before anything runs in production.\n\n\u003C!-- ------------------------ --\u003E\n\n## Conclusion And Resources\nDuration: 2\n\nYou now have Dataiku Cobuild connected to Snowflake Cortex and a production AI pipeline built entirely through natural language. From here you can apply the same workflow to new projects, refactor existing Dataiku flows, or use Cobuild to generate documentation for pipelines your team already owns.\n\n### What You Learned\n\n- How to configure Snowflake for use with Dataiku Cobuild\n- How to connect Snowflake Cortex to a Dataiku instance\n- How to build and deploy a production AI pipeline using natural language in Cobuild\n\n### Related Resources\n\n- [Dataiku Cobuild on Snowflake](https://www.dataiku.com/product/cobuild-on-snowflake)\n- [Dataiku Cobuild Knowledge Base](https://knowledge.dataiku.com/latest/ui/cobuild/concept-cobuild.html)\n- [Dataiku Documentation](https://doc.dataiku.com)\n- [Snowflake Cortex REST API](https://docs.snowflake.com/en/user-guide/snowflake-cortex/cortex-rest-api)\n- [Snowflake Cortex AI Documentation](https://docs.snowflake.com/en/guides-overview-ai-features)\n- [Dataiku on Snowflake Marketplace](https://app.snowflake.com/marketplace/listing/GZTSZNUFX99I/dataiku-dataiku-for-enterprise-ai)\n","multiValue":false,":type":"text/x-markdown"},"quickstartArticleLogoImage":{"dataType":"string","title":"Quickstart Article Logo Image","multiValue":false,":type":"text/plain"}},"isDeveloperGuidesPage":false,"elementsOrder":["quickstartArticleBody","quickstartArticleLogoImage"],":type":"snowflake-site/components/contentfragment","model":"snowflake-site/models/quickstart-article"},"flexible_column_cont":{"id":"flexible-column-container-0b29e90682","type":"2-column-75-25","alignColumns":"top","containerMaxWidth":"extra-large","topPadding":"none","bottomPadding":"none","spaceBetween":"none","reverseOnMobile":false,"carouselOnMobile":false,"backgroundImageOption":"none","flexible_column_content_container_1":{"layout":"SIMPLE","id":"container-535afd3bbc",":items":{"quickstart_last_modi":{"id":"quickstart-last-modified-a8382cf573","icon":{"id":"icon","icon":"calendar",":type":"snowflake-site/components/icon","appliedCssClassNames":"snowflake-icon-blue"},"lastModifiedDatePrefix":"Updated","lastModifiedDate":"2026-08-07",":type":"snowflake-site/components/quickstart/quickstart-last-modified","appliedCssClassNames":"snowflake-responsive-component-top-padding-small"},"text":{"id":"text-5c991b1904","additionalClasses":"qs-disclaimer-text","text":"\u003Cp\u003E\u003Cspan style=\"color: #666;\"\u003EThis content is provided as is, and is not maintained on an ongoing basis. It may be out of date with current Snowflake instances\u003C/span\u003E\u003C/p\u003E\r\n","richText":true,":type":"snowflake-site/components/text","appliedCssClassNames":"snowflake-responsive-component-top-padding-small"}},":itemsOrder":["quickstart_last_modi","text"],":type":"snowflake-site/components/flexible-column-container/flexible-column-content-container"},"flexible_column_content_container_2":{"layout":"SIMPLE","id":"container-361fec05df",":items":{},":itemsOrder":[],":type":"snowflake-site/components/flexible-column-container/flexible-column-content-container"},"isBlogPage":false,"isActiveTOC":false,":type":"snowflake-site/components/flexible-column-container"}},":itemsOrder":["contentfragment","flexible_column_cont"],":type":"snowflake-site/components/flexible-column-container/flexible-column-content-container"},"flexible_column_content_container_2":{"layout":"SIMPLE","id":"container-18f7d71463",":items":{"quickstart_table_of_":{"layout":"SIMPLE","id":"container-8d409f9b5c","isDeveloperGuidesPage":false,":items":{"quickstart_table_of_":{"id":"quickstart-table-of-content-916569d4bb","fragmentPath":"/content/dam/snowflake-site/en/content-fragments/quickstarts/build-production-ai-with-dataiku-cobuild-on-snowflake",":type":"snowflake-site/components/quickstart/quickstart-table-of-content","headings":["\u003Ch2\u003EOverview\u003C/h2\u003E","\u003Ch2\u003ESnowflake Setup Steps\u003C/h2\u003E","\u003Ch2\u003EConnecting Cortex to Dataiku\u003C/h2\u003E","\u003Ch2\u003EUsing Dataiku Cobuild\u003C/h2\u003E","\u003Ch2\u003EConclusion And Resources\u003C/h2\u003E"]},"quickstart_button":{"id":"quickstart-button-f26fcc4574","fragmentPath":"/content/dam/snowflake-site/en/content-fragments/quickstarts/build-production-ai-with-dataiku-cobuild-on-snowflake",":type":"snowflake-site/components/quickstart/quickstart-button","appliedCssClassNames":"snowflake-responsive-component-top-padding-none"}},":itemsOrder":["quickstart_table_of_","quickstart_button"],":type":"snowflake-site/components/quickstart/quickstart-table-of-content/quickstart-table-of-content-container"}},":itemsOrder":["quickstart_table_of_"],":type":"snowflake-site/components/flexible-column-container/flexible-column-content-container"},"isBlogPage":false,"isActiveTOC":false,":type":"snowflake-site/components/flexible-column-container"},"markup_editor":{"id":"markup-editor-92df3f5c4d","title":"Page CSS","cssContent":"#quickstart-template-main-flexible-container{padding:24px}#quickstart-template-main-flexible-container \u003E .snowflake-flexible-column-container-items{grid-template-columns:1fr 0}.qs-disclaimer-text p \u003E span{font-size:15px !important}@media (min-width:768px){#quickstart-template-main-flexible-container{padding:24px 32px}#quickstart-template-main-flexible-container \u003E .snowflake-flexible-column-container-items{grid-template-columns:7fr 3fr;gap:48px}}@media (max-width:767px){#quickstart-template-main-flexible-container \u003E .snowflake-flexible-column-container-items{gap:0}}@media (min-width:1024px){#quickstart-template-main-flexible-container{padding:0 92px 48px 92px}#quickstart-template-main-flexible-container \u003E .snowflake-flexible-column-container-items{gap:117px}}","isGSAPEnabled":false,":type":"snowflake-site/components/markup-editor"}},":itemsOrder":["quickstart_hero","flexible_column_cont","markup_editor"],":type":"wcm/foundation/components/responsivegrid"},"modal_container":{"layout":"SIMPLE","id":"container-5fe7279134",":items":{},":itemsOrder":[],":type":"snowflake-site/components/modal/modal-container"},"experiencefragment-footer":{"id":"experiencefragment-1c971c0eb3","localizedFragmentVariationPath":"/content/experience-fragments/snowflake-site/language-masters/en/site/footer/master/jcr:content","configured":true,"xfModelPath":"/content/experience-fragments/snowflake-site/language-masters/en/site/footer/master.xfmodel.json",":type":"snowflake-site/components/experiencefragment"},"markup_editor":{"id":"markup-editor-0acab7d329","title":"Quickstarts Overrides","cssContent":".snowflake-markdown blockquote{padding:24px 32px;background:#f6f9fa;border:1px solid #29b5e8;border-radius:16px}.snowflake-markdown .snowflake-image-container img{width:auto !important;max-width:100%}.snowflake-markdown .snowflake-text ol{padding-left:20px !important}.snowflake-markdown .snowflake-text li{margin:0 0 12px 0 !important}.snowflake-markdown h3.snowflake-markdown-h3{font-size:20px !important;font-family:Texta,sans-serif !important}@media (min-width:768px){.snowflake-markdown h3.snowflake-markdown-h3{font-size:28px !important}}","isGSAPEnabled":false,":type":"snowflake-site/components/markup-editor"}},":itemsOrder":["experiencefragment-banner","experiencefragment-header","markup_editor_1950346551","responsivegrid","modal_container","experiencefragment-footer","markup_editor"],":type":"wcm/foundation/components/responsivegrid"}},":itemsOrder":["root"],":hierarchyType":"page",":path":"/content/snowflake-site/global/en/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake","isPasswordProtected":false,"analyticsContentTags":["snowflake-site:taxonomy/solution-center/certification/quickstart","snowflake-site:taxonomy/product/ai","snowflake-site:taxonomy/snowflake-feature/model-development","snowflake-site:taxonomy/solution-center/certification/partner-solution","snowflake-site:taxonomy/snowflake-feature/applied-analytics"],"analyticsEnabled":true,":mappedPath":"/en/developers/guides/build-production-ai-with-dataiku-cobuild-on-snowflake/",":type":"snowflake-site/components/structure/page","locale":"en"}