--- title: From Shadows to Spotlight - How Swiss Post Performs... description: Swiss Post's solution architect describes the IRIS project, a fraud detection ML system deployed in production using shadow model deployment. The architecture... canonical: https://daily.dev/posts/from-shadows-to-spotlight---how-swiss-post-performs-reliable-ml-deployment-by-giovanni-degiorgi-uw5nbtltl twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: From Shadows to Spotlight - How Swiss Post Performs Reliable ML Deployment by Giovanni Degiorgi | daily.dev og:description: Swiss Post's solution architect describes the IRIS project, a fraud detection ML system deployed in production using shadow model deployment. The architecture... og:url: https://daily.dev/posts/from-shadows-to-spotlight---how-swiss-post-performs-reliable-ml-deployment-by-giovanni-degiorgi-uw5nbtltl og:image: https://api.daily.dev/og/posts/Uw5NBtLtL.png og:image:alt: From Shadows to Spotlight - How Swiss Post Performs Reliable ML Deployment by Giovanni Degiorgi og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # From Shadows to Spotlight - How Swiss Post Performs Reliable ML Deployment by Giovanni Degiorgi **[Devoxx](https://daily.dev/sources/devoxx)** · 42 min read · 0 upvotes · 0 comments ## Summary Swiss Post's solution architect describes the IRIS project, a fraud detection ML system deployed in production using shadow model deployment. The architecture uses three separate AWS accounts for training, inference, and serving, with SageMaker pipelines automating model training on production data stored in governed data rooms. The core technique is shadow deployment: routing live traffic to both a production model and a shadow model simultaneously, storing both outputs in S3, then comparing operational metrics and model drift (data drift, concept drift, prediction drift, score distribution drift) to decide whether to promote the shadow model. A Lambda function automates the promotion decision based on drift metrics and synthetic test data with known labels. Key outcomes: zero-impact production testing, automated model promotion only when performance improves, and low infrastructure cost. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Similar posts on daily.dev - [\[AWS-AI\] Amazon SageMaker Explained: From Data Preparation to Model Deployment](https://daily.dev/posts/aws-ai-amazon-sagemaker-explained-from-data-preparation-to-model-deployment-izl4ihjhw) · Medium · 0 upvotes · 0 comments --- Tags: [#machine-learning](https://daily.dev/tags/machine-learning), [#mlops](https://daily.dev/tags/mlops), [#fraud-detection](https://daily.dev/tags/fraud-detection), [#aws-sagemaker](https://daily.dev/tags/aws-sagemaker) [View this post on daily.dev](https://daily.dev/posts/from-shadows-to-spotlight---how-swiss-post-performs-reliable-ml-deployment-by-giovanni-degiorgi-uw5nbtltl) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"From Shadows to Spotlight - How Swiss Post Performs Reliable ML Deployment by Giovanni Degiorgi","url":"https://daily.dev/posts/from-shadows-to-spotlight---how-swiss-post-performs-reliable-ml-deployment-by-giovanni-degiorgi-uw5nbtltl","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/from-shadows-to-spotlight---how-swiss-post-performs-reliable-ml-deployment-by-giovanni-degiorgi-uw5nbtltl"},"datePublished":"2026-04-01T08:25:18.466Z","dateModified":"2026-04-01T08:26:14.486Z","description":"Swiss Post's solution architect describes the IRIS project, a fraud detection ML system deployed in production using shadow model deployment. The architecture...","image":"https://i.ytimg.com/vi/jygFpQ9hao4/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/jygFpQ9hao4/sddefault.jpg","isAccessibleForFree":true,"articleSection":"Devoxx","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Devoxx","logo":"https://media.daily.dev/image/upload/s--6vG797wn--/f_auto,q_auto/v1773648800/logos/devoxx?_a=BAMAMiiu0","url":"https://daily.dev/sources/devoxx"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/from-shadows-to-spotlight---how-swiss-post-performs-reliable-ml-deployment-by-giovanni-degiorgi-uw5nbtltl","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"machine-learning,mlops,fraud-detection,aws-sagemaker","timeRequired":"PT42M","video":{"@type":"VideoObject","name":"From Shadows to Spotlight - How Swiss Post Performs Reliable ML Deployment by Giovanni Degiorgi","description":"Swiss Post's solution architect describes the IRIS project, a fraud detection ML system deployed in production using shadow model deployment. The architecture...","thumbnailUrl":"https://i.ytimg.com/vi/jygFpQ9hao4/sddefault.jpg","uploadDate":"2026-04-01T08:25:18.466Z","duration":"PT42M","url":"https://api.daily.dev/r/Uw5NBtLtL","embedUrl":"https://www.youtube.com/embed/jygFpQ9hao4"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Devoxx","item":"https://daily.dev/sources/devoxx"},{"@type":"ListItem","position":3,"name":"From Shadows to Spotlight - How Swiss Post Performs Reliable ML Deployment by Giovanni Degiorgi"}]} ```