{"id":12032,"date":"2026-11-19T08:30:00","date_gmt":"2026-11-19T08:30:00","guid":{"rendered":"https:\/\/www.cotocus.com\/blog\/?p=12032"},"modified":"2026-11-19T08:30:00","modified_gmt":"2026-11-19T08:30:00","slug":"sketching-ml-pipeline-plain-language-explainer","status":"publish","type":"post","link":"https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/","title":{"rendered":"Sketching an ML Pipeline on Paper: A Plain-Language Guide"},"content":{"rendered":"<p>Machine learning projects have a reputation for sprawling. A notebook becomes a script, the script becomes a scheduled job, and a year later nobody can say where the training data comes from or why the model behaves differently on Mondays. A sketch of the ML pipeline, drawn on paper before any of that happens, is an unglamorous but dependable defence against that drift. This explainer describes what such a sketch is, which boxes belong on it and how to hunt for weak spots, using everyday language rather than jargon.<\/p>\n<h2>What is a pipeline sketch, really?<\/h2>\n<p>A pipeline sketch is a rough drawing of how information travels through a machine learning system, from the moment raw data is collected to the moment a prediction reaches a user. Each stage is a box. Each handover is an arrow. That is all it is. No notation to learn, no special software, no right answer. If a colleague can look at your page and retell the journey of a single data record, the sketch is doing its job.<\/p>\n<h2>Why draw it before building?<\/h2>\n<p>Because a pipeline has many more joins than most people expect, and every join is a place where assumptions hide. Does the training data match what the live system will see? Who notices if the incoming data suddenly changes shape? Where does the model file live, and how does the serving code pick up a new version? Writing code reveals these questions slowly and expensively. A drawing reveals them quickly and for free. The point is not accuracy; it is early discovery.<\/p>\n<p>There is also an argument from evidence. Teams that adopt machine learning in production commonly report that the model itself is a small part of the work, and that data handling, deployment and monitoring take the bulk of the effort. A sketch gives those larger parts visible space on the page, rather than leaving them as footnotes to a clever algorithm.<\/p>\n<h2>The five boxes to draw first<\/h2>\n<ol>\n<li><strong>Where the data comes from.<\/strong> Name the real sources: a transactions table, a sensor feed, a folder of documents. Note who owns each one.<\/li>\n<li><strong>How it is prepared.<\/strong> Draw the cleaning, joining and reshaping steps as one or two boxes, and write beside them what could go wrong, such as missing values or changed column meanings.<\/li>\n<li><strong>How the model is trained and checked.<\/strong> Show where experiments are recorded and what test decides whether a new model is good enough to replace the old one.<\/li>\n<li><strong>How it is released.<\/strong> Sketch the path from an approved model to the place where it answers requests, including how to go back to a previous version.<\/li>\n<li><strong>How it is watched.<\/strong> Draw what you will observe once it is live: input data, predictions, and eventual real outcomes.<\/li>\n<\/ol>\n<p>For a structured overview of the stages a team typically moves through as it matures, the <a href=\"https:\/\/www.cotocus.com\/blog\/mlops-roadmap-a-complete-guide-for-data-scientists-and-engineering-teams\/\">step-by-step MLOps route for data scientists and engineers<\/a> can be read alongside your sketch to check whether any stage is missing.<\/p>\n<h2>How to read your own sketch critically<\/h2>\n<p>Once the boxes are down, interrogate the arrows. Ask three questions of each.<\/p>\n<ul>\n<li><strong>What travels along it?<\/strong> If you cannot say whether it is a file, a table or a live request, the interface is undefined.<\/li>\n<li><strong>How often?<\/strong> Daily batches, hourly refreshes and instant calls have different needs, and mixing them unconsciously is a classic source of surprises.<\/li>\n<li><strong>What if it fails?<\/strong> If the answer is &#8220;we would not know&#8221;, draw a small eye beside the arrow and plan a way to look.<\/li>\n<\/ul>\n<p>Then step back and look for lopsidedness. A page that is crowded on the left with data preparation and nearly empty on the right where monitoring should be is a typical pattern, and it predicts the kind of project that works at launch and quietly degrades afterwards.<\/p>\n<h2>A small worked example<\/h2>\n<p>Imagine a team forecasting weekly demand for a regional retailer. Their first sketch shows sales data going into a model and a forecast coming out. Reading it critically, they ask where the sales data is corrected for returns, and realise nobody has said. They add a box. They ask what happens in a week when a store closes and the feed goes quiet; the answer is that the model would predict as if nothing had changed. They add an arrow from a simple checks box. They ask how they will know the forecasts are slipping, and add a comparison of predictions against real sales. In twenty minutes, the sketch has grown from two boxes to six, and each new box is a problem solved before it became an incident. The broader case for running models with the same care as any other production service is made in <a href=\"https:\/\/www.cotocus.com\/blog\/the-essential-role-of-mlops-in-machine-learning-success\/\">this piece on the part MLOps plays when models must succeed<\/a>.<\/p>\n<h2>Two sketches are better than one<\/h2>\n<p>A useful refinement is to draw the pipeline twice: once for how the system will behave on a normal day, and once for the day something changes. The second drawing might show a source that starts sending a new format, a retraining job that runs late or a model that is rolled back. Placing the two pages side by side makes it obvious which boxes have a plan for change and which assume that nothing ever moves. Teams are often surprised that the &#8220;ordinary day&#8221; page is tidy while the &#8220;changed day&#8221; page is full of question marks, and that contrast tells them where to spend their next hour of design effort.<\/p>\n<p>It also helps to put a rough owner beside each box, using roles rather than names, such as &#8220;data team&#8221; or &#8220;platform engineers&#8221;. Boxes with no owner are boxes that will be orphaned later.<\/p>\n<h2>Keeping the sketch honest<\/h2>\n<p>The drawing should stay rough for as long as the design is uncertain. Polishing it is a signal that the team has stopped questioning it. Pin it where people will see it, and invite anyone to add a sticky note with a worry. When the build begins, the sketch becomes a reference for what you intended, and when something goes wrong later it is quicker to compare reality against a simple picture than against a sprawling codebase.<\/p>\n<h2>Support material outside the pipeline<\/h2>\n<p>Work like this often happens across several devices, and the odd hardware hiccup is inevitable during a workshop. For quick answers about everyday gadgets, <a href=\"https:\/\/osnivo.com\/\">Osnivo<\/a> is an independent reference built around device families, with checkers and a community forum, and it can save a facilitator a few minutes without turning the session into a support call.<\/p>\n<h2>The short version<\/h2>\n<p>A pipeline sketch is a cheap way to learn where a machine learning system is vague before it is expensive to fix. Draw the journey of one data record, name every handover, ask what happens when each one fails and check that watching the system gets as much space on the page as building it. Then, and only then, start writing code that fits the picture.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>A plain-language explainer on drawing a machine learning pipeline before building it, with five boxes to start and questions for every arrow.<\/p>\n","protected":false},"author":37,"featured_media":12033,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[147],"tags":[],"class_list":["post-12032","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-devops"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v28.6 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Sketching an ML Pipeline on Paper: A Plain-Language Guide - Cotocus<\/title>\n<meta name=\"description\" content=\"A plain-language guide to sketching an ML pipeline on paper: five boxes to draw first and how to read your drawing for weak spots before coding.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Sketching an ML Pipeline on Paper: A Plain-Language Guide - Cotocus\" \/>\n<meta property=\"og:description\" content=\"A plain-language guide to sketching an ML pipeline on paper: five boxes to draw first and how to read your drawing for weak spots before coding.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/\" \/>\n<meta property=\"og:site_name\" content=\"Cotocus\" \/>\n<meta property=\"article:published_time\" content=\"2026-11-19T08:30:00+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/www.cotocus.com\/blog\/wp-content\/uploads\/2026\/11\/b2a-banner-3363598-banner.png\" \/>\n\t<meta property=\"og:image:width\" content=\"1200\" \/>\n\t<meta property=\"og:image:height\" content=\"630\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"Nikhil Farooqui\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Nikhil Farooqui\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"6 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/sketching-ml-pipeline-plain-language-explainer\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/sketching-ml-pipeline-plain-language-explainer\\\/\"},\"author\":{\"name\":\"Nikhil Farooqui\",\"@id\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/#\\\/schema\\\/person\\\/419007bef27e8d85890550d783b96c50\"},\"headline\":\"Sketching an ML Pipeline on Paper: A Plain-Language Guide\",\"datePublished\":\"2026-11-19T08:30:00+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/sketching-ml-pipeline-plain-language-explainer\\\/\"},\"wordCount\":1167,\"commentCount\":0,\"image\":{\"@id\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/sketching-ml-pipeline-plain-language-explainer\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/wp-content\\\/uploads\\\/2026\\\/11\\\/b2a-banner-3363598-banner.png\",\"articleSection\":[\"DevOps\"],\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/sketching-ml-pipeline-plain-language-explainer\\\/#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/sketching-ml-pipeline-plain-language-explainer\\\/\",\"url\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/sketching-ml-pipeline-plain-language-explainer\\\/\",\"name\":\"Sketching an ML Pipeline on Paper: A Plain-Language Guide - Cotocus\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/sketching-ml-pipeline-plain-language-explainer\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/sketching-ml-pipeline-plain-language-explainer\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/wp-content\\\/uploads\\\/2026\\\/11\\\/b2a-banner-3363598-banner.png\",\"datePublished\":\"2026-11-19T08:30:00+00:00\",\"author\":{\"@id\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/#\\\/schema\\\/person\\\/419007bef27e8d85890550d783b96c50\"},\"description\":\"A plain-language guide to sketching an ML pipeline on paper: five boxes to draw first and how to read your drawing for weak spots before coding.\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/sketching-ml-pipeline-plain-language-explainer\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/sketching-ml-pipeline-plain-language-explainer\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/sketching-ml-pipeline-plain-language-explainer\\\/#primaryimage\",\"url\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/wp-content\\\/uploads\\\/2026\\\/11\\\/b2a-banner-3363598-banner.png\",\"contentUrl\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/wp-content\\\/uploads\\\/2026\\\/11\\\/b2a-banner-3363598-banner.png\",\"width\":1200,\"height\":630},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/sketching-ml-pipeline-plain-language-explainer\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Sketching an ML Pipeline on Paper: A Plain-Language Guide\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/#website\",\"url\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/\",\"name\":\"Cotocus\",\"description\":\"Shaping Tomorrow\u2019s Tech Today\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/#\\\/schema\\\/person\\\/419007bef27e8d85890550d783b96c50\",\"name\":\"Nikhil Farooqui\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/d855702dac00a5af0405bb329101be495718d611233e52c86fe360c8fca6f6dc?s=96&d=mm&r=g\",\"url\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/d855702dac00a5af0405bb329101be495718d611233e52c86fe360c8fca6f6dc?s=96&d=mm&r=g\",\"contentUrl\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/d855702dac00a5af0405bb329101be495718d611233e52c86fe360c8fca6f6dc?s=96&d=mm&r=g\",\"caption\":\"Nikhil Farooqui\"},\"url\":\"https:\\\/\\\/www.cotocus.com\\\/blog\\\/author\\\/nikhil-farooqui\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Sketching an ML Pipeline on Paper: A Plain-Language Guide - Cotocus","description":"A plain-language guide to sketching an ML pipeline on paper: five boxes to draw first and how to read your drawing for weak spots before coding.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/","og_locale":"en_US","og_type":"article","og_title":"Sketching an ML Pipeline on Paper: A Plain-Language Guide - Cotocus","og_description":"A plain-language guide to sketching an ML pipeline on paper: five boxes to draw first and how to read your drawing for weak spots before coding.","og_url":"https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/","og_site_name":"Cotocus","article_published_time":"2026-11-19T08:30:00+00:00","og_image":[{"width":1200,"height":630,"url":"https:\/\/www.cotocus.com\/blog\/wp-content\/uploads\/2026\/11\/b2a-banner-3363598-banner.png","type":"image\/png"}],"author":"Nikhil Farooqui","twitter_card":"summary_large_image","twitter_misc":{"Written by":"Nikhil Farooqui","Est. reading time":"6 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/#article","isPartOf":{"@id":"https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/"},"author":{"name":"Nikhil Farooqui","@id":"https:\/\/www.cotocus.com\/blog\/#\/schema\/person\/419007bef27e8d85890550d783b96c50"},"headline":"Sketching an ML Pipeline on Paper: A Plain-Language Guide","datePublished":"2026-11-19T08:30:00+00:00","mainEntityOfPage":{"@id":"https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/"},"wordCount":1167,"commentCount":0,"image":{"@id":"https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/#primaryimage"},"thumbnailUrl":"https:\/\/www.cotocus.com\/blog\/wp-content\/uploads\/2026\/11\/b2a-banner-3363598-banner.png","articleSection":["DevOps"],"inLanguage":"en-US","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/#respond"]}]},{"@type":"WebPage","@id":"https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/","url":"https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/","name":"Sketching an ML Pipeline on Paper: A Plain-Language Guide - Cotocus","isPartOf":{"@id":"https:\/\/www.cotocus.com\/blog\/#website"},"primaryImageOfPage":{"@id":"https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/#primaryimage"},"image":{"@id":"https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/#primaryimage"},"thumbnailUrl":"https:\/\/www.cotocus.com\/blog\/wp-content\/uploads\/2026\/11\/b2a-banner-3363598-banner.png","datePublished":"2026-11-19T08:30:00+00:00","author":{"@id":"https:\/\/www.cotocus.com\/blog\/#\/schema\/person\/419007bef27e8d85890550d783b96c50"},"description":"A plain-language guide to sketching an ML pipeline on paper: five boxes to draw first and how to read your drawing for weak spots before coding.","breadcrumb":{"@id":"https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/#primaryimage","url":"https:\/\/www.cotocus.com\/blog\/wp-content\/uploads\/2026\/11\/b2a-banner-3363598-banner.png","contentUrl":"https:\/\/www.cotocus.com\/blog\/wp-content\/uploads\/2026\/11\/b2a-banner-3363598-banner.png","width":1200,"height":630},{"@type":"BreadcrumbList","@id":"https:\/\/www.cotocus.com\/blog\/sketching-ml-pipeline-plain-language-explainer\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.cotocus.com\/blog\/"},{"@type":"ListItem","position":2,"name":"Sketching an ML Pipeline on Paper: A Plain-Language Guide"}]},{"@type":"WebSite","@id":"https:\/\/www.cotocus.com\/blog\/#website","url":"https:\/\/www.cotocus.com\/blog\/","name":"Cotocus","description":"Shaping Tomorrow\u2019s Tech Today","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.cotocus.com\/blog\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Person","@id":"https:\/\/www.cotocus.com\/blog\/#\/schema\/person\/419007bef27e8d85890550d783b96c50","name":"Nikhil Farooqui","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/secure.gravatar.com\/avatar\/d855702dac00a5af0405bb329101be495718d611233e52c86fe360c8fca6f6dc?s=96&d=mm&r=g","url":"https:\/\/secure.gravatar.com\/avatar\/d855702dac00a5af0405bb329101be495718d611233e52c86fe360c8fca6f6dc?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/d855702dac00a5af0405bb329101be495718d611233e52c86fe360c8fca6f6dc?s=96&d=mm&r=g","caption":"Nikhil Farooqui"},"url":"https:\/\/www.cotocus.com\/blog\/author\/nikhil-farooqui\/"}]}},"_links":{"self":[{"href":"https:\/\/www.cotocus.com\/blog\/wp-json\/wp\/v2\/posts\/12032","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.cotocus.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.cotocus.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.cotocus.com\/blog\/wp-json\/wp\/v2\/users\/37"}],"replies":[{"embeddable":true,"href":"https:\/\/www.cotocus.com\/blog\/wp-json\/wp\/v2\/comments?post=12032"}],"version-history":[{"count":1,"href":"https:\/\/www.cotocus.com\/blog\/wp-json\/wp\/v2\/posts\/12032\/revisions"}],"predecessor-version":[{"id":12139,"href":"https:\/\/www.cotocus.com\/blog\/wp-json\/wp\/v2\/posts\/12032\/revisions\/12139"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.cotocus.com\/blog\/wp-json\/wp\/v2\/media\/12033"}],"wp:attachment":[{"href":"https:\/\/www.cotocus.com\/blog\/wp-json\/wp\/v2\/media?parent=12032"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.cotocus.com\/blog\/wp-json\/wp\/v2\/categories?post=12032"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.cotocus.com\/blog\/wp-json\/wp\/v2\/tags?post=12032"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}