{"id":43,"date":"2026-05-18T12:37:25","date_gmt":"2026-05-18T12:37:25","guid":{"rendered":"https:\/\/convly.ai\/neural-networks-explained\/"},"modified":"2026-06-15T18:18:20","modified_gmt":"2026-06-15T18:18:20","slug":"neural-networks-explained","status":"publish","type":"post","link":"https:\/\/convly.ai\/pt\/neural-networks-explained\/","title":{"rendered":"Redes neurais explicadas para n\u00e3o engenheiros (guia 2026)"},"content":{"rendered":"<p>Neural networks are the engine behind modern AI \u2014 every chatbot, image generator, and voice assistant runs on them. The name sounds intimidating, and most explanations drown you in math. They don&#8217;t have to. The core idea behind a neural network is genuinely understandable without any equations. This guide explains it clearly, for non-engineers.<\/p>\n<div class=\"convly-tldr\">\n<h3>Principais conclus\u00f5es<\/h3>\n<ul>\n<li><strong>A neural network<\/strong> is a system of simple connected units that together learn complex patterns.<\/li>\n<li><strong>It&#8217;s loosely inspired by the brain<\/strong> \u2014 but it&#8217;s math, not biology.<\/li>\n<li><strong>It learns by adjusting &#8220;weights&#8221;<\/strong> \u2014 connection strengths \u2014 to reduce its errors.<\/li>\n<li><strong>Layers build understanding<\/strong> \u2014 early layers catch simple features, later layers combine them into complex ones.<\/li>\n<li><strong>&#8220;Deep learning&#8221;<\/strong> just means a neural network with many layers.<\/li>\n<\/ul>\n<\/div>\n<div id=\"ez-toc-container\" class=\"ez-toc-v2_0_84 counter-flat ez-toc-counter ez-toc-container-direction\">\n<label for=\"ez-toc-cssicon-toggle-item-6a38a9392c052\" class=\"ez-toc-cssicon-toggle-label\"><span class=\"\"><span class=\"eztoc-hide\" style=\"display:none;\">Alternar<\/span><span class=\"ez-toc-icon-toggle-span\"><svg style=\"fill: #000000;color:#000000\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" class=\"list-377408\" width=\"20px\" height=\"20px\" viewbox=\"0 0 24 24\" fill=\"none\"><path d=\"M6 6H4v2h2V6zm14 0H8v2h12V6zM4 11h2v2H4v-2zm16 0H8v2h12v-2zM4 16h2v2H4v-2zm16 0H8v2h12v-2z\" fill=\"currentColor\"><\/path><\/svg><svg style=\"fill: #000000;color:#000000\" class=\"arrow-unsorted-368013\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" width=\"10px\" height=\"10px\" viewbox=\"0 0 24 24\" version=\"1.2\" baseprofile=\"tiny\"><path d=\"M18.2 9.3l-6.2-6.3-6.2 6.3c-.2.2-.3.4-.3.7s.1.5.3.7c.2.2.4.3.7.3h11c.3 0 .5-.1.7-.3.2-.2.3-.5.3-.7s-.1-.5-.3-.7zM5.8 14.7l6.2 6.3 6.2-6.3c.2-.2.3-.5.3-.7s-.1-.5-.3-.7c-.2-.2-.4-.3-.7-.3h-11c-.3 0-.5.1-.7.3-.2.2-.3.5-.3.7s.1.5.3.7z\"\/><\/svg><\/span><\/span><\/label><input type=\"checkbox\"  id=\"ez-toc-cssicon-toggle-item-6a38a9392c052\"  aria-label=\"Alternar\" \/><nav><ul class='ez-toc-list ez-toc-list-level-1' ><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-1\" href=\"https:\/\/convly.ai\/pt\/neural-networks-explained\/#What_is_a_neural_network\" >What is a neural network?<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-2\" href=\"https:\/\/convly.ai\/pt\/neural-networks-explained\/#The_building_block_an_artificial_neuron\" >The building block: an artificial neuron<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-3\" href=\"https:\/\/convly.ai\/pt\/neural-networks-explained\/#Layers_how_the_network_is_organized\" >Layers: how the network is organized<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-4\" href=\"https:\/\/convly.ai\/pt\/neural-networks-explained\/#How_a_neural_network_learns\" >How a neural network learns<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-5\" href=\"https:\/\/convly.ai\/pt\/neural-networks-explained\/#A_simple_analogy\" >A simple analogy<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-6\" href=\"https:\/\/convly.ai\/pt\/neural-networks-explained\/#Main_types_of_neural_networks\" >Main types of neural networks<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-7\" href=\"https:\/\/convly.ai\/pt\/neural-networks-explained\/#Neural_networks_and_deep_learning\" >Neural networks and deep learning<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-8\" href=\"https:\/\/convly.ai\/pt\/neural-networks-explained\/#How_to_start_building_your_first_neural_network\" >How to start building your first neural network<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-9\" href=\"https:\/\/convly.ai\/pt\/neural-networks-explained\/#FAQ\" >Perguntas frequentes<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-10\" href=\"https:\/\/convly.ai\/pt\/neural-networks-explained\/#Bottom_line\" >Conclus\u00e3o<\/a><\/li><li class='ez-toc-page-1'><a class=\"ez-toc-link ez-toc-heading-11\" href=\"https:\/\/convly.ai\/pt\/neural-networks-explained\/#Related_articles\" >Artigos relacionados<\/a><\/li><\/ul><\/nav><\/div>\n<h2><span class=\"ez-toc-section\" id=\"What_is_a_neural_network\"><\/span>What is a neural network?<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>A neural network is a method for finding patterns in data, built from many small, simple parts working together. Each part \u2014 a &#8220;neuron&#8221; \u2014 does something trivial on its own. But connect thousands or millions of them in layers, and the whole network can do remarkable things: recognize faces, translate languages, generate text.<\/p>\n<p>The name comes from a loose inspiration: the human brain is a network of connected neurons. But don&#8217;t take the analogy too far. An artificial neural network is not a digital brain \u2014 it&#8217;s a mathematical structure that happens to share one organizing idea with biology: <strong>many simple units, richly connected, produce complex behavior.<\/strong><\/p>\n<h2><span class=\"ez-toc-section\" id=\"The_building_block_an_artificial_neuron\"><\/span>The building block: an artificial neuron<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Strip it down and a single artificial neuron does three things:<\/p>\n<ol>\n<li><strong>Receives inputs<\/strong> \u2014 numbers coming from the data, or from other neurons.<\/li>\n<li><strong>Weighs them<\/strong> \u2014 each input is multiplied by a &#8220;weight,&#8221; a number that says how important that input is. The neuron adds the weighted inputs together.<\/li>\n<li><strong>Decides an output<\/strong> \u2014 it passes that sum through a simple function that decides what number to send onward.<\/li>\n<\/ol>\n<p>That&#8217;s it. One neuron is almost too simple to be useful. The power comes entirely from connecting many of them.<\/p>\n<h2><span class=\"ez-toc-section\" id=\"Layers_how_the_network_is_organized\"><\/span>Layers: how the network is organized<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Neurons are arranged in <strong>layers<\/strong>, and data flows through them in order:<\/p>\n<ul>\n<li><strong>The input layer<\/strong> receives the raw data. For an image, this might be the pixel values; for text, the words converted to numbers.<\/li>\n<li><strong>The hidden layers<\/strong> are the middle layers where the real work happens. Each one transforms the data a little, passing its result to the next.<\/li>\n<li><strong>The output layer<\/strong> produces the final answer \u2014 a category, a probability, a predicted number, the next word.<\/li>\n<\/ul>\n<p>The crucial insight is what the hidden layers do in sequence. In an image network, the <strong>first hidden layer<\/strong> might learn to detect simple things \u2014 edges and patches of color. The <strong>next layer<\/strong> combines edges into shapes \u2014 corners, curves. A <strong>later layer<\/strong> combines shapes into parts \u2014 an eye, a wheel. The <strong>final layers<\/strong> combine parts into whole concepts \u2014 a face, a car.<\/p>\n<p>Each layer builds on the one before, turning simple features into complex understanding. That layered build-up is the secret of how neural networks handle messy, real-world data.<\/p>\n<h2><span class=\"ez-toc-section\" id=\"How_a_neural_network_learns\"><\/span>How a neural network learns<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>A fresh neural network is useless \u2014 its weights are random, so its outputs are random. Learning is the process of finding good weights. It works as a loop:<\/p>\n<ol>\n<li><strong>Make a prediction.<\/strong> Feed in a training example and let the network produce an output.<\/li>\n<li><strong>Measure the error.<\/strong> Compare the output to the known correct answer. The gap is the error (often called the &#8220;loss&#8221;).<\/li>\n<li><strong>Assign blame.<\/strong> Work backwards through the network to figure out how much each weight contributed to the error. This step is called <strong>backpropagation<\/strong>.<\/li>\n<li><strong>Adjust the weights.<\/strong> Nudge every weight slightly in the direction that would have reduced the error.<\/li>\n<li><strong>Repeat.<\/strong> Do this across thousands or millions of examples, many times over.<\/li>\n<\/ol>\n<p>Each pass makes the network a tiny bit better. After enough passes, the weights settle into values that capture the real pattern \u2014 and the network can handle new inputs it never saw. That cycle of <em>predict, measure, blame, adjust<\/em> is the entire essence of training.<\/p>\n<h2><span class=\"ez-toc-section\" id=\"A_simple_analogy\"><\/span>A simple analogy<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Imagine tuning a huge mixing board with thousands of sliders, trying to produce a perfect sound. You play a note, hear how far off it is, and adjust each slider a little. You can&#8217;t get it right in one move \u2014 but with enough small, guided adjustments, the sound converges on what you want.<\/p>\n<p>A neural network is that mixing board. The sliders are the weights. The &#8220;how far off&#8221; is the error. And training is the patient, automated process of making millions of tiny, guided adjustments until the output is right.<\/p>\n<h2><span class=\"ez-toc-section\" id=\"Main_types_of_neural_networks\"><\/span>Main types of neural networks<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Different problems use different network designs:<\/p>\n<table class=\"convly-vs\">\n<thead>\n<tr>\n<th>Tipo<\/th>\n<th>Good at<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Feedforward networks<\/td>\n<td>Basic prediction and classification<\/td>\n<\/tr>\n<tr>\n<td>Convolutional networks (CNNs)<\/td>\n<td>Images and computer vision<\/td>\n<\/tr>\n<tr>\n<td>Recurrent networks (RNNs)<\/td>\n<td>Sequences \u2014 older approach for text and time series<\/td>\n<\/tr>\n<tr>\n<td>Transformers<\/td>\n<td>Language and beyond \u2014 the architecture behind modern AI<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>O <strong>transformer<\/strong> is the one that matters most today. It&#8217;s the architecture behind large language models, modern image generators, and most of the AI breakthroughs of recent years. Its key trick is &#8220;attention&#8221; \u2014 the ability to weigh which parts of the input matter most for each part of the output.<\/p>\n<h2><span class=\"ez-toc-section\" id=\"Neural_networks_and_deep_learning\"><\/span>Neural networks and deep learning<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>You&#8217;ll often hear &#8220;deep learning&#8221; alongside neural networks. The relationship is simple: <strong>deep learning means using neural networks with many hidden layers<\/strong> (&#8220;deep&#8221; = many layers). Early networks had one or two hidden layers; modern ones can have dozens or far more. More layers let the network learn richer, more abstract patterns \u2014 which is why deep learning unlocked the current AI era. Our <a href=\"\/pt\/deep-learning-vs-machine-learning\/\">deep learning vs machine learning guide<\/a> covers this further.<\/p>\n<p><!--ai-enriched--><\/p>\n<h2><span class=\"ez-toc-section\" id=\"How_to_start_building_your_first_neural_network\"><\/span>How to start building your first neural network<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Understanding the theory is one thing; training a working network is another. The good news is that you no longer write neural networks from scratch. Mature open-source frameworks handle the heavy math \u2014 the gradients, the matrix multiplications, the GPU acceleration \u2014 so you can define a model in a few lines of Python and focus on learning how the pieces fit together.<\/p>\n<p>Three frameworks dominate in 2026, and the right one depends on where you are:<\/p>\n<ul>\n<li><strong>Keras<\/strong> \u2014 the gentlest on-ramp. Its high-level API lets you stack layers with minimal boilerplate, so you can train a real classifier on your first afternoon. It runs on top of TensorFlow (and now other backends), which makes it ideal for learning concepts before you worry about internals.<\/li>\n<li><strong>PyTorch<\/strong> \u2014 the researcher&#8217;s default and, by most measures, the most-used framework in published deep-learning work. Its eager, Pythonic style behaves like normal code, so debugging is straightforward and almost every new model or tutorial you find online is written in it. This is the one to grow into.<\/li>\n<li><strong>TensorFlow<\/strong> \u2014 still a backbone of large-scale production deployment, with strong tooling for serving models on phones, browsers, and servers. Most beginners reach for it via Keras rather than directly.<\/li>\n<\/ul>\n<p>A practical first project is image classification on a small built-in dataset such as MNIST handwritten digits or Fashion-MNIST. They are tiny, ship with every framework, and train on a laptop in minutes \u2014 no GPU required. Building one teaches the full loop: load data, define layers, pick a <strong>loss function<\/strong>, train across several passes (epochs), then check accuracy on data the model never saw.<\/p>\n<p>You also do not need to buy hardware to begin. Free cloud notebooks such as Google Colab and Kaggle Kernels give you a GPU in the browser, which is more than enough for early experiments. A dedicated GPU only starts to matter once you train larger models or your own image and text datasets.<\/p>\n<p>A sensible path: start in Keras to build intuition, reproduce a tutorial end to end, then rewrite the same model in PyTorch to understand what the high-level calls were doing. Once the training loop feels familiar, move from toy datasets to a problem you actually care about \u2014 that is where the learning compounds fastest.<\/p>\n<h2><span class=\"ez-toc-section\" id=\"FAQ\"><\/span>Perguntas frequentes<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<h3>What is a neural network in simple terms?<\/h3>\n<p>A neural network is a pattern-finding system built from many simple connected units called neurons, arranged in layers. Each neuron does a tiny calculation; together, the layers transform raw data into a useful answer. It learns by adjusting the connection strengths between neurons to reduce its mistakes.<\/p>\n<h3>Are neural networks like the human brain?<\/h3>\n<p>Only loosely. They borrow one idea from biology \u2014 many simple units connected together produce complex behavior \u2014 but an artificial neural network is a mathematical structure, not a digital brain. It doesn&#8217;t think or understand the way a brain does.<\/p>\n<h3>What is the difference between a neural network and deep learning?<\/h3>\n<p>A neural network is the structure. Deep learning is the practice of using neural networks with many layers (&#8220;deep&#8221; networks). All deep learning uses neural networks, and deep networks are what made modern AI possible.<\/p>\n<h3>How do neural networks learn?<\/h3>\n<p>Through a loop: the network makes a prediction, measures how wrong it was, uses backpropagation to find which weights caused the error, and adjusts those weights slightly to reduce it. Repeating this across huge amounts of data, many times, gradually produces an accurate network.<\/p>\n<h3>What are weights in a neural network?<\/h3>\n<p>Weights are numbers that set the strength of each connection between neurons. They determine how much influence one neuron&#8217;s output has on the next. Learning is essentially the process of finding the right values for all the weights \u2014 that&#8217;s where the network&#8217;s &#8220;knowledge&#8221; is stored.<\/p>\n<h3>Do I need to be good at math to build a neural network?<\/h3>\n<p>To use a framework like Keras or PyTorch, no \u2014 you can train a working model knowing only basic Python. Modern libraries handle the calculus and linear algebra for you. That said, a working feel for the underlying ideas pays off quickly: comfort with vectors and matrices helps you reason about layer shapes, and a rough grasp of derivatives makes the training process (and why it sometimes fails) far less mysterious. You can pick this math up alongside the code rather than before it.<\/p>\n<h3>How long does it take to learn neural networks?<\/h3>\n<p>You can train your first working model in an afternoon by following a tutorial. Reaching the point where you can build a network for your own problem, diagnose why it is underperforming, and tune it sensibly typically takes a few months of consistent, hands-on practice. The fastest route is to build small projects end to end rather than only watching courses \u2014 debugging your own broken model teaches more than any lecture.<\/p>\n<h3>Can I learn to build neural networks for free?<\/h3>\n<p>Yes. The major frameworks are open source, the standard beginner datasets ship with them, and free cloud notebooks such as Google Colab and Kaggle provide a GPU in the browser at no cost. Combined with the extensive free documentation and tutorials each framework publishes, you can go from zero to a trained model without spending anything or buying special hardware.<\/p>\n<h2><span class=\"ez-toc-section\" id=\"Bottom_line\"><\/span>Conclus\u00e3o<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>A neural network is not magic and not a brain. It&#8217;s a layered structure of simple units that learns by making predictions, measuring its errors, and adjusting millions of internal weights until it gets things right. Layers stacked on layers turn simple features into complex understanding \u2014 and &#8220;deep learning&#8221; is just this idea taken to many layers.<\/p>\n<p>That single mechanism \u2014 <em>predict, measure, adjust, repeat<\/em> \u2014 powers nearly every AI system you use. Understand that loop and you understand the foundation of modern artificial intelligence. To see how it fits the bigger picture, start with <a href=\"\/pt\/what-is-machine-learning-beginners-guide\/\">what machine learning is<\/a>.<\/p>\n<p><!--related-block--><\/p>\n<div class=\"convly-related\">\n<h2><span class=\"ez-toc-section\" id=\"Related_articles\"><\/span>Artigos relacionados<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<ul>\n<li><a href=\"https:\/\/convly.ai\/pt\/what-is-a-vector-database-2026\/\">O que \u00e9 um banco de dados vetorial? (Guia 2026)<\/a><\/li>\n<li><a href=\"https:\/\/convly.ai\/pt\/overfitting-how-to-prevent-it\/\">Sobreajuste no aprendizado de m\u00e1quina: o que \u00e9 e como evit\u00e1-lo<\/a><\/li>\n<li><a href=\"https:\/\/convly.ai\/pt\/best-free-datasets-machine-learning\/\">15 Best Free Datasets for Machine Learning Projects (2026)<\/a><\/li>\n<li><a href=\"https:\/\/convly.ai\/pt\/deep-learning-vs-machine-learning\/\">Aprendizado profundo versus aprendizado de m\u00e1quina: as principais diferen\u00e7as (2026)<\/a><\/li>\n<li><a href=\"https:\/\/convly.ai\/pt\/top-10-machine-learning-algorithms\/\">Top 10 Machine Learning Algorithms Every Beginner Should Know<\/a><\/li>\n<\/ul>\n<\/div>","protected":false},"excerpt":{"rendered":"<p>What is a neural network, really? A clear, no-math explanation of how neural networks work \u2014 neurons, layers, training \u2014 for anyone without an engineering background.<\/p>","protected":false},"author":0,"featured_media":44,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"default","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","ast-disable-related-posts":"","theme-transparent-header-meta":"","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"default","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"footnotes":""},"categories":[2],"tags":[466,459,464,463,465],"class_list":["post-43","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-machine-learning","tag-ai-explained","tag-deep-learning","tag-how-neural-networks-work","tag-neural-networks","tag-neurons"],"_links":{"self":[{"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/posts\/43","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/comments?post=43"}],"version-history":[{"count":4,"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/posts\/43\/revisions"}],"predecessor-version":[{"id":1154,"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/posts\/43\/revisions\/1154"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/media\/44"}],"wp:attachment":[{"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/media?parent=43"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/categories?post=43"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/convly.ai\/pt\/wp-json\/wp\/v2\/tags?post=43"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}