{"version":1,"pages":[{"id":"b0wEwqrjXkOpxDBbxNWp","title":"Overview","pathname":"/mlops-playbook","siteSpaceId":"sitesp_Jwu8f","description":""},{"id":"uYJz9hPjOHYkzZHxXD0w","title":"Key terms","pathname":"/mlops-playbook/overview/key-terms","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Overview"}]},{"id":"B0KxCIDfyN6QFSAYOBoz","title":"What is MLOps","pathname":"/mlops-playbook/what-is-mlops","siteSpaceId":"sitesp_Jwu8f","description":""},{"id":"5sSs2TaQneIYN0qf16Uu","title":"Principles","pathname":"/mlops-playbook/principles","siteSpaceId":"sitesp_Jwu8f","description":""},{"id":"DZYS7s79YPAFcVfEuMN6","title":"Solid data foundations","pathname":"/mlops-playbook/principles/solid-data-foundations","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Principles"}]},{"id":"24vgMtlIpGaTmzy6ezT3","title":"Provide an environment that allows data scientists to create and test models","pathname":"/mlops-playbook/principles/provide-an-environment-that-allows-data-scientists-to-create-and-test-models","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Principles"}]},{"id":"ma9No3y1hHYleBL0QsBU","title":"A machine learning  service is  a product","pathname":"/mlops-playbook/principles/a-machine-learning-service-is-a-product","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Principles"}]},{"id":"jn0y32blLESO5SMEsOuD","title":"Apply continuous delivery","pathname":"/mlops-playbook/principles/apply-continuous-delivery","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Principles"}]},{"id":"alfxXeGzhGouRPKwUaU2","title":"Evaluate and monitor  algorithms throughout their lifecycle","pathname":"/mlops-playbook/principles/evaluate-and-monitor-algorithms-throughout-their-lifecycle","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Principles"}]},{"id":"ZbaaW7kqDLG2kGEn46Ph","title":"MLOps is a team effort","pathname":"/mlops-playbook/principles/mlops-is-a-team-effort","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Principles"}]},{"id":"aAKCJT3ScLSEtwJft02Z","title":"Practices","pathname":"/mlops-playbook/practices","siteSpaceId":"sitesp_Jwu8f","description":""},{"id":"pAjZRPUKhjj9uDv022ZV","title":"Collect performance data","pathname":"/mlops-playbook/practices/collect-performance-data","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Practices"}]},{"id":"SgX8kWxdT91bBLkdx5En","title":"Ways of deploying your model","pathname":"/mlops-playbook/practices/ways-of-deploying-your-model","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Practices"}]},{"id":"6VkFWo7YmPO5uukH5z9s","title":"How often do you deploy a model?","pathname":"/mlops-playbook/practices/how-often-do-you-deploy-a-model","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Practices"}]},{"id":"aU0BdmllEDurkTTasxGf","title":"Keep a versioned model repository","pathname":"/mlops-playbook/practices/keep-a-versioned-model-repository","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Practices"}]},{"id":"cOVNKiD3huniL1U7lP7W","title":"Measure and proactively evaluate quality of training data","pathname":"/mlops-playbook/practices/measure-and-proactively-evaluate-quality-of-training-data","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Practices"}]},{"id":"wJmHJJQYgxEGNNZvLhjc","title":"Testing through the ML pipeline","pathname":"/mlops-playbook/practices/testing-through-the-ml-pipeline","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Practices"}]},{"id":"VgrukHvBGt124DVMfRAe","title":"Business impact is more than just accuracy - understand your baseline","pathname":"/mlops-playbook/practices/business-impact-is-more-than-just-accuracy-understand-your-baseline","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Practices"}]},{"id":"BmRlSkBtCH022VQrhKhs","title":"Regularly monitor your model in production","pathname":"/mlops-playbook/practices/regularly-monitor-your-model-in-production","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Practices"}]},{"id":"lwDv77XVtfKMQP92RGk2","title":"Monitor data quality","pathname":"/mlops-playbook/practices/monitor-data-quality","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Practices"}]},{"id":"pCrOGys5je8k1mu2jsP2","title":"Automate the model lifecycle","pathname":"/mlops-playbook/practices/automate-the-model-lifecycle","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Practices"}]},{"id":"RrQmRoYSofhnvj9EWlaW","title":"Create a walking skeleton/steel thread","pathname":"/mlops-playbook/practices/create-a-walking-skeleton-steel-thread","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Practices"}]},{"id":"H8UkFvYt9K5ofG44t1E5","title":"Appropriately optimise models for inference","pathname":"/mlops-playbook/practices/appropriately-optimise-models-for-inference","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Practices"}]},{"id":"38Z2AlY9b8ryFs1fsEn6","title":"Explore","pathname":"/mlops-playbook/explore","siteSpaceId":"sitesp_Jwu8f","description":""},{"id":"VBS0xBqupENW2koF3MS3","title":"Pitfalls (Avoid)","pathname":"/mlops-playbook/pitfalls-avoid","siteSpaceId":"sitesp_Jwu8f","description":""},{"id":"X0iMmjMXW0PRnbbnadeu","title":"User Trust and Engagement","pathname":"/mlops-playbook/pitfalls-avoid/user-trust-and-engagement","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Pitfalls (Avoid)"}]},{"id":"MVACYUOzHflWSde47kau","title":"Explainability","pathname":"/mlops-playbook/pitfalls-avoid/explainability","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Pitfalls (Avoid)"}]},{"id":"ByueBAKwqE5oRM5fzBfD","title":"Avoid  notebooks in production","pathname":"/mlops-playbook/pitfalls-avoid/avoid-notebooks-in-production","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Pitfalls (Avoid)"}]},{"id":"yR3j8YfqaJfNNSnZ7Ef5","title":"Poor security practices","pathname":"/mlops-playbook/pitfalls-avoid/poor-security-practices","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Pitfalls (Avoid)"}]},{"id":"poLonVjefy793EG8bOga","title":"Don’t treat accuracy as the only or even the best way to  evaluate your algorithm","pathname":"/mlops-playbook/pitfalls-avoid/dont-treat-accuracy-as-the-only-or-even-the-best-way-to-evaluate-your-algorithm","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Pitfalls (Avoid)"}]},{"id":"LqR6RgwEA0NUzzj5x6iu","title":"Use machine learning judiciously","pathname":"/mlops-playbook/pitfalls-avoid/use-machine-learning-judiciously","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Pitfalls (Avoid)"}]},{"id":"kIdBe2dispW5zOsHQVwI","title":"Don’t forget to understand the at-inference usage profile","pathname":"/mlops-playbook/pitfalls-avoid/dont-forget-to-understand-the-at-inference-usage-profile","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Pitfalls (Avoid)"}]},{"id":"KlkD5mr01w5DKbCqPdVw","title":"Don’t make it difficult for a data scientists to access data or use the tools they need","pathname":"/mlops-playbook/pitfalls-avoid/dont-make-it-difficult-for-a-data-scientists-to-access-data-or-use-the-tools-they-need","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Pitfalls (Avoid)"}]},{"id":"ULaWpKdlVLYJSPE0P0vI","title":"Not taking into consideration the downstream application of the model","pathname":"/mlops-playbook/pitfalls-avoid/not-taking-into-consideration-the-downstream-application-of-the-model","siteSpaceId":"sitesp_Jwu8f","description":"","breadcrumbs":[{"label":"Pitfalls (Avoid)"}]},{"id":"b3NpQyiCyfBnrsCO2qmR","title":"Contributors","pathname":"/mlops-playbook/contributors","siteSpaceId":"sitesp_Jwu8f","description":""}]}