AI-102 Microsoft Azure AI Solution Complete Exam Prep

Install Python and PyCharm to set up coding environments for cognitive services, explore language options, IDE choices, and access sample code and repositories for Azure AI skills.

Explore Azure AI services categorized into language, speech, vision, Applied AI (decision) and OpenAI, and learn how to choose one or combine multiple categories for a given solution.

Explore Azure AI Vision services and Custom Vision to analyze images with tags, captions, OCR, metadata, and train models with labeled images via Vision Studio.

Explore Azure AI Language services, merging Text Analytics, Language Understanding, and QnA Maker, and use Language Studio to extract key phrases and summarize text.

Azure AI decision services include anomaly detector, content safety, and personalizer; anomaly detector and personalizer are deprecated, while content safety remains active with a content safety studio for moderation.

Explore Azure's speech services, including speech to text and text to speech, real-time translation, and speaker recognition, with hands-on demos in Speech Studio and practical resource setup.

Explore how to start with an azure free account, access cognitive services, and choose between multi-service and individual accounts in the azure portal.

Create a computer vision resource in the Azure portal, choose East US 2 for Vision Studio, and use the endpoint and security token to analyze images.

Learn how the cognitive service endpoint uses keys-based authentication with two rotatable keys you can regenerate, testable via code to analyze images and view confidence scores for detected features.

Set up diagnostics and monitoring for cognitive services, create alert rules for latency, errors, and key regeneration, and use action groups to email or text when alerts fire.

monitor cognitive services through metrics in the Azure portal, tracking total calls, errors, data in, data out, latency, and availability for the computer vision service.

Configure diagnostic settings to collect audit and request logs and metrics for cognitive services, store them in a log analytics workspace, and query with Kusto to analyze performance.

Create and configure a multi-service cognitive services resource in Azure, test it with keys and endpoints, and prepare for cost analysis across computer vision and other APIs.

Learn to manage Azure cognitive services costs by analyzing pricing per region and per transaction, understand cost reports, invoices, and budgeting with forecasts and alerts.

Learn to manage cognitive services endpoints and keys, and secure access with selected networks, ip filters, and private endpoints, including a custom domain for speech service.

Explore how cognitive search uses Azure Active Directory authentication with role-based access control to assign roles like owner or search service contributor, and compare keys with role-based access control.

Explore the responsible AI principles guiding Azure cognitive services, including fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability.

Identify the default endpoint for a service by recognizing current domains such as cognitiveservices.azure.com, search.windows.net, and cognitive.microsoft.com, and note that older services use regional endpoints like eastus2.api.cognitive.microsoft.com.

Configure CI/CD for an Azure bot web app by linking a GitHub repository to auto update on code commits, noting Azure cognitive service deployment is normally not required.

Deploy azure cognitive services in containers to run anomaly detector, text analytics, key phrase extraction, and sentiment analysis locally with a docker image, reducing cloud dependency.

Create an anomaly detector in Azure AI Services to analyze time series data, using univariate or multivariate detection, with streaming or batch options, authenticated via endpoint and API key.

Explore anomaly detector testing with a Microsoft test harness, demonstrating univariate streaming and bulk modes, adjusting sensitivity to identify anomalies across univariate and multivariate data.

Create and configure the Azure content moderation service, then build a .NET app that uses text and image moderation APIs to detect profanity, PII, and faces, with list management.

Explore how Azure Personalizer uses reinforcement learning with context, actions, and rewards to improve recommendations, including apprentice and online modes, via rank and reward APIs.

Learn to implement azure personalizer in a .NET console app by configuring context and actions, and using the rank API with rewards to determine recommendations.

Explore metrics advisor, a service built on anomaly detector that learns from time series data with no data science background, ingests data from SQL databases, and alerts via a dashboard.

Discover Azure Immersive Reader, a text reader with focus mode and features like parts of speech and a picture dictionary, and learn how to embed it via HTML and JavaScript.

Explore implementing image and video processing with Computer Vision Service, Azure Form Recognizer, Custom Vision Service, and Azure Video Indexer to meet ai-102 exam objectives.

Demonstrate extracting image tags with the Azure computer vision API, creating a client with endpoint and key, and using tag image for remote or local images.

Use the describe image method in Azure cognitive services to retrieve image descriptions for remote URLs and local images, producing captions with confidence scores.

Explore domain-specific models in Microsoft Azure Vision API to identify landmarks and celebrities, and compare results with generic tagging by analyzing remote and local images.

Learn to detect brands and logos in images using the computer vision API, returning brand names, confidence, and logo locations with x,y coordinates.

Explore how the computer vision service analyzes images for adult, racy, and gorry content, returning booleans for each category and a confidence score to guide moderation.

Generate thumbnails from an image using an ai service that crops to the area of interest, focusing on faces or landmarks, by specifying width, height, and the image url.

Create .NET Core 3.1 console app in Visual Studio that uses Azure cognitive services computer vision to analyze an image, authenticates with endpoint and key, and retrieves captions and features.

Learn how the computer vision API extracts text from images using OCR for printed text. Compare handwritten text with ink recognizer and Reed command asynchronous results.

Use the azure form recognizer APIs to extract data from forms and receipts, including vendor details, items, totals, addresses, and dates, then export to a database.

Apply handwriting OCR with Visual Studio 2019 and C# to read handwriting from an Azure blob image using the read file method and extract text.

Learn to detect faces and extract attributes using the Azure Face API, assign face IDs, and locate matching faces across images.

Create a person group, train the model with your own images, and use the Azure Face API to identify individuals in new test images.

Call the face API via HTTP to extract age, emotion, gender, glasses, makeup, and other attributes, then examine the returned JSON.

Build a Visual Studio 2019 C# console app to use the face API for detecting faces, mapping up to 27 landmarks, and estimating age, gender, emotion, smile, and head pose.