{"id":735469,"date":"2021-03-22T20:21:04","date_gmt":"2021-03-23T03:21:04","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-project&#038;p=735469"},"modified":"2024-03-27T18:07:47","modified_gmt":"2024-03-28T01:07:47","slug":"seagull","status":"publish","type":"msr-project","link":"https:\/\/www.microsoft.com\/en-us\/research\/project\/seagull\/","title":{"rendered":"Reusable AI Infrastructure"},"content":{"rendered":"<section class=\"mb-3 moray-highlight\">\n\t<div class=\"card-img-overlay mx-lg-0\">\n\t\t<div class=\"card-background  has-background-catalina-blue card-background--full-bleed\">\n\t\t\t\t\t<\/div>\n\t\t<!-- Foreground -->\n\t\t<div class=\"card-foreground d-flex mt-md-n5 my-lg-5 px-g px-lg-0\">\n\t\t\t<!-- Container -->\n\t\t\t<div class=\"container d-flex mt-md-n5 my-lg-5 align-self-center\">\n\t\t\t\t<!-- Card wrapper -->\n\t\t\t\t<div class=\"w-100 w-lg-col-5\">\n\t\t\t\t\t<!-- Card -->\n\t\t\t\t\t<div class=\"card material-md-card py-5 px-md-5\">\n\t\t\t\t\t\t<div class=\"card-body \">\n\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\n\n<h1 class=\"wp-block-heading h2\" id=\"reusable-ai-infrastructure\">Reusable AI Infrastructure<\/h1>\n\n\t\t\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t<\/div>\n\t\t<\/div>\n\t<\/div>\n<\/section>\n\n\n\n\n\n<p>We propose an end-to-end reusable AI infrastructure that automates the whole life-cycle of AI-driven applications from data ingestion and validation, feature engineering, choice of predictive technique, training, deployment of ML models to a REST endpoint, versioning of deployed models, inference, and evaluation of KPI metrics. The infrastructure leverages Azure Machine Learning and the big data platform Cosmos.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"status\">Status<\/h3>\n\n\n\n<p>Deployed WW to optimize backup scheduling for PostgreSQL and MySQL servers<\/p>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"impact\">Impact<\/h3>\n\n\n\n<ol>\n<li>A reusable AI infrastructure reduces the engineering effort to incorporate AI capabilities into cloud applications from months to weeks. <\/li>\n\n\n\n<li>Our solution avoids backup collisions with peaks of customer activity and therefore achieves several hundred hours of improved customer experience per month in all Azure regions.<\/li>\n<\/ol>\n\n\n\n<figure class=\"wp-block-image aligncenter is-resized\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"732\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2021\/03\/infra-1024x732.png\" alt=\"Seagull: graphic showing case-agnostic offline components\" class=\"wp-image-742165\" style=\"width:534px;height:381px\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2021\/03\/infra-1024x732.png 1024w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2021\/03\/infra-300x214.png 300w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2021\/03\/infra-768x549.png 768w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2021\/03\/infra-1536x1098.png 1536w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2021\/03\/infra-16x12.png 16w, https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2021\/03\/infra.png 1845w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><figcaption class=\"wp-element-caption\">Reusable AI Infrastructure<\/figcaption><\/figure>\n\n\n<span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span>","protected":false},"excerpt":{"rendered":"<p>We propose an end-to-end reusable AI infrastructure that automates the whole life-cycle of AI-driven applications from data ingestion and validation, feature engineering, choice of predictive technique, training, deployment of ML models to a REST endpoint, versioning of deployed models, inference, and evaluation of KPI metrics. The infrastructure leverages Azure Machine Learning and the big data [&hellip;]<\/p>\n","protected":false},"featured_media":735604,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"footnotes":""},"research-area":[13563],"msr-impact-theme":[],"msr-pillar":[],"msr_project_start":"2018-01-29","related-publications":[723718],"related-downloads":[],"related-videos":[],"related-groups":[],"related-events":[],"related-opportunities":[],"related-posts":[],"related-articles":[],"tab-content":[],"slides":[],"related-researchers":[{"type":"user_nicename","display_name":"Olga Poppe","user_id":40054,"people_section":"Section name 0","alias":"olpoppe"}],"msr_research_lab":[],"msr_impact_theme":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/735469"}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-project"}],"version-history":[{"count":54,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/735469\/revisions"}],"predecessor-version":[{"id":1018932,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/735469\/revisions\/1018932"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media\/735604"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=735469"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=735469"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=735469"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=735469"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}