AI-900: MS Azure AI Fundamentals – Learn & Get Certified

Discover how machine learning teaches computers from past data to make predictions about future outcomes. Explore concepts like features, labels, training, and how supervised and unsupervised methods differ.

Machines learn from data to find patterns and make predictions. Supervised learning uses labeled inputs and answers (regression and classification), while unsupervised learning discovers patterns through clustering.

Learn how regression uses historical data to predict continuous numbers, such as ice cream sales from temperature, by fitting a regression line. Assess accuracy with hold-out data.

Compare binary and multi-class classification in supervised learning, using probabilities from a logistic function and a confusion matrix to measure accuracy, precision, and recall, with emails and penguin species examples.

Explore clustering as unsupervised learning that groups data into natural clusters without labels, using k-means and moving centroids. Evaluate cluster quality with the silhouette score.

Explore neural networks and deep learning, from artificial neurons and layered architectures to training with backpropagation and gradient descent, learning to minimize loss across epochs.

Transformers power modern AI through attention, enabling parallel processing and global context understanding for text, images, speech, and beyond, fueling tools like ChatGPT and Copilot.

Learn how to avail a free 200 USD credit by creating a free Azure account, going through the setup steps, and understanding pay-as-you-go considerations.

Explore generative AI, a branch of AI that creates new content—text, images, code, and more—driven by large language models and neural networks.

Explore how language models power modern AI systems, including chatbots and translation apps, by tokenizing text, mapping tokens to embeddings, and using transformers with attention to handle long text.

Explore how transformers leverage self-attention, multi-head attention, and positional encoding to scale with parallel processing, powering large language models, transfer learning, and multimodal applications.

Not all language models are the same; large language models and small language models differ in size, training, and use cases, from cloud powered generalists to locally run domain specialists.

Master prompting techniques to elicit high quality AI answers by being specific, structuring requests, using roles, step-by-step guidance, and iterative refinement.

Explore responsible generative AI across text, image creation, answering questions, and learning-assistant roles. Apply fairness, transparency, reliability, privacy, and accountability to design, test, and use AI with human oversight.

Explore a Microsoft demo of a generative AI chat assistant that explains expense limits, submits claims, and answers hotel and meal queries, with clear policy information.

See a two-step demo where you select a candidate's resume, choose a job opening, and watch an AI-generated score reflect skill fit.

Create a resource group in the Azure portal, then add a resource and a project for generative AI using the Azure AI Foundry, and explore the Azure AI Foundry Portal.

Explore generative AI in the Azure AI portal by creating a deployment in the chat playground, selecting base models, and crafting instructions to tailor a travel destination finder.

Explore how natural language processing enables computers to read, understand, and respond to human language through language understanding and generation, powering chatbots, translators, voice assistants, and Azure AI Language Services.

Explore how machines tokenize text, identify parts of speech, map syntax and semantics, and use embeddings to infer meaning, context, intent, and sentiment in language.

Explore the shift from counting words to understanding meaning through semantics and word embeddings, enabling deep learning models like GPT and BERT. Apply this to sentiment analysis, translation, and chatbots.

See natural language processing in action as an audio-to-text portal converts audio to transcript and automatically extracts language, translation, named entities, sentiment, and key phrases using the Azure portal.

Explore natural language processing in the Azure portal by creating an AI project, using the language playground to extract named entities, summarize text, detect language, translate, and analyze sentiment.

Discover how speech AI enables computers to perform speech recognition and speech synthesis, turning voice into text and text into natural speech, enabling assistants, transcription, and accessibility.

Explore how speech recognition turns voice into text and speech synthesis creates natural voice output, using acoustic and language models, phonemes, and prosody for real-time, human-like AI speech.

See how text-to-speech and speech-to-text work in AI services, converting voice to text and back to speech, with an Azure portal demo.

Explore text to speech and speech to text in the Azure AI portal, with real time transcription, language auto-detect, diarization, and a voice gallery to convert text to speech.

Discover how computer vision lets machines see and interpret the visual world from images and videos. Explore applications like image analysis, facial recognition, and optical character recognition in real-world AI.

Explore how image processing, through pixel-by-pixel analysis, filters, and convolution, yields feature maps that prepare images for convolutional neural networks to recognize objects.

Explore how machine learning powers vision by training models on labeled images to classify objects, from feature extraction to deep learning, and apply these ideas to real-world vision tasks.

Discover how modern vision models advance computer vision from CNNs to transformers, using transfer learning on pre-trained models for healthcare, retail, and autonomous vehicles, with Azure AI Vision.

Learn how image analysis works in Azure AI: enable image analysis in the portal, upload an image, and see the model identify a soccer ball using image identification methods.

Explore image captioning, dense captioning, tag extraction, and object detection in the Azure AI portal. Create a hub, configure AI services, and read captions with bounding boxes across images.

Extract information from unstructured data with AI-powered information extraction, OCR, and multimodal techniques to create structured data. Explore invoices and resumes, and see how Azure AI enables fast, accurate extraction.

Explore how artificial intelligence extracts data from images, forms, and mixed documents using vision, form, and multimodal extraction. Learn how knowledge mining connects these outputs into a searchable knowledge base.

Explore how information extraction highlights invoice fields like vendor name, address, and total amount using optical character recognition on uploaded images. The demo previews Azure portal integration.

See content understanding in the Azure portal in action, from creating a project to invoice data extraction, post-call analysis, and viewing extracted fields.