Introduction to AI

What is AI?

  • AI is the creation of software that imitates human behaviours and capabilities.

Key workloads of AI

  • Machine learning - "Teaching" a computer model to make predictions and draw conclusions from data.

  • Anomaly detection - Automatically detecting errors or unusual activity in a system.

  • Computer vision - Interpreting the world visually through cameras, video, and images.

  • Natural language processing - Interpreting written or spoken language, and responding in kind.

  • Knowledge mining - Extracting information from large volumes of often unstructured data to create a searchable knowledge store.

Understand machine learning

How does machine learning work?

  • Machines learn from data.

  • Huge volumes of data are generated in our everyday lives.

  • Data scientists use this data to train machine-learning models.

  • Models make predictions and inferences based on the relationships found in the data.

A real-world example of machine learning in sustainable farming

  • The Yield, an agricultural technology company in Australia, uses sensors, data, and machine learning to help farmers make informed decisions.

  • Machine learning is used to analyze weather, soil, and plant conditions.

Machine learning in Microsoft Azure

  • Azure Machine Learning service is a cloud-based platform for creating, managing, and publishing machine learning models.

  • Provides features and capabilities for machine learning.

Automated machine learningThis feature enables non-experts to quickly create an effective machine-learning model from data.
Azure Machine Learning designerA graphical interface enabling no-code development of machine learning solutions.
Data and compute managementCloud-based data storage and compute resources that professional data scientists can use to run data experiment code at scale.
PipelinesData scientists, software engineers, and IT operations professionals can define pipelines to orchestrate model training, deployment, and management tasks.

Understand anomaly detection

Scenario Examples

  • Credit card transaction monitoring to detect fraud

  • Automated production line activity tracking to identify failures

  • Racing car telemetry system to proactively warn about mechanical failures

Anomaly Detection

  • Anomaly detection is a machine learning-based technique

  • It analyzes data over time and identifies unusual changes

Anomaly Detection in Microsoft Azure

  • Microsoft Azure Anomaly Detector service offers an API for developers

  • Developers can use this API to create anomaly detection solutions.

A race car drives past and an instrument panel shows telemetry values, which vary over time. When an anomaly occurs, a warning is displayed and the car stops.

  1. Sensors in the car collect telemetry, such as engine revolutions, brake temperature, and so on.

  2. An anomaly detection model is trained to understand expected fluctuations in the telemetry measurements over time.

  3. If a measurement occurs outside of the normal expected range, the model reports an anomaly that can be used to alert the race engineer to call the driver in for a pit stop to fix the issue before it forces retirement from the race.

Understand computer vision

Computer Vision is an area of AI that deals with visual processing.

Let's explore some of the possibilities that computer vision brings.

The Seeing AI app is a great example of the power of computer vision. Designed for the blind and low-vision community, the Seeing AI app harnesses the power of AI to open up the visual world and describe nearby people, text and objects.

Computer Vision models and capabilities

Most computer vision solutions are based on machine learning models that can be applied to visual input from cameras, videos, or images.

The following table describes common computer vision tasks.

Image classificationImage classification involves training a machine learning model to classify images based on their contents. For example, in a traffic monitoring solution you might use an image classification model to classify images based on the type of vehicle they contain, such as taxis, buses, cyclists, and so on.

An image of a taxi with the label "Taxi"

Object detection

Object detection machine learning models are trained to classify individual objects within an image, and identify their location with a bounding box. For example, a traffic monitoring solution might use object detection to identify the location of different classes of vehicle.

An image of a street with buses, cars, and cyclists identified and highlighted with a bounding box

Semantic segmentation

Semantic segmentation is an advanced machine learning technique in which individual pixels in the image are classified according to the object to which they belong. For example, a traffic monitoring solution might overlay traffic images with "mask" layers to highlight different vehicles using specific colors.

An image of a street with the pixels belonging to buses, cars, and cyclists identified

Image analysis

You can create solutions that combine machine learning models with advanced image analysis techniques to extract information from images, including "tags" that could help catalog the image or even descriptive captions that summarize the scene shown in the image.

An image of a person with a dog on a street and the caption "A person with a dog on a street"

Computer vision services in Microsoft Azure

Azure AI VisionYou can use this service to analyze images and video, and extract descriptions, tags, objects, and text.
Azure AI Custom VisionUse this service to train custom image classification and object detection models using your own images.
Azure AI FaceThe Azure AI Face service enables you to build face detection and facial recognition solutions.
Azure AI Document IntelligenceUse this service to extract information from scanned forms and documents.

Natural language processing

Definition of NLP

  • Natural language processing (NLP) is the area of AI that deals with creating software that understands written and spoken language

Capabilities of NLP

  • Analyze and interpret text in documents, email messages, and other sources

  • Interpret spoken language, and synthesize speech responses

  • Automatically translate spoken or written phrases between languages

  • Interpret commands and determine appropriate actions

Example of NLP in a VR game

  • Starship Commander is a virtual reality (VR) game that uses NLP

  • Players can control the narrative and interact with in-game characters and starship systems using NLP

Natural language processing in Microsoft Azure

Azure AI LanguageUse this service to access features for understanding and analyzing text, training language models that can understand spoken or text-based commands, and building intelligent applications.
Azure AI TranslatorUse this service to translate text between more than 60 languages.
Azure AI SpeechUse this service to recognize and synthesize speech, and to translate spoken languages.
Azure AI Bot ServiceThis service provides a platform for conversational AI, the capability of a software "agent" to participate in a conversation. Developers can use the Bot Framework to create a bot and manage it with Azure Bot Service - integrating back-end services like Language, and connecting to channels for web chat, email, Microsoft Teams, and others.

Knowledge Mining

Knowledge mining

  • Knowledge mining involves extracting information from large volumes of unstructured data

  • Creates a searchable knowledge store

Azure Cognitive Search

  • A knowledge mining solution in Microsoft Azure

  • Private, enterprise search solution with tools for building indexes

  • Indexes can be used for internal or public-facing content

Azure Cognitive Search and AI

  • Utilizes AI capabilities of Azure AI services

  • Includes image processing, content extraction, and natural language processing

  • Can index previously unsearchable documents

  • Extracts and surfaces insights from large amounts of data quickly.

Challenges and risks with AI

Artificial Intelligence is a powerful tool that can be used to greatly benefit the world. However, like any tool, it must be used responsibly.

Challenge or RiskExample
Bias can affect resultsA loan-approval model discriminates by gender due to bias in the data with which it was trained
Errors may cause harmAn autonomous vehicle experiences a system failure and causes a collision
Data could be exposedA medical diagnostic bot is trained using sensitive patient data, which is stored insecurely
Solutions may not work for everyoneA home automation assistant provides no audio output for visually impaired users
Users must trust a complex systemAn AI-based financial tool makes investment recommendations - what are they based on?

Responsible AI

Treating all people fairly

  • AI systems should treat all people fairly

  • Avoid bias based on gender, ethnicity, or other factors that result in an unfair advantage or disadvantage

Azure Machine Learning and fairness

  • Azure Machine Learning includes the capability to interpret models and quantify the extent of each feature's influence on predictions

  • This helps identify and mitigate bias in the model

Responsible AI and facial recognition

  • Microsoft's implementation of Responsible AI includes retiring facial recognition capabilities

  • These capabilities can be misused to infer emotional states and identity attributes

  • Subjecting people to stereotyping, discrimination, or unfair denial of services

Reliability and Safety

Importance of reliability and safety in AI systems

  • AI systems in critical applications like autonomous vehicles or medical diagnoses must perform reliably and safely

  • Unreliable AI systems can pose a significant risk to human life

Testing and deployment management for AI software

  • AI-based software applications must undergo rigorous testing and deployment management processes

  • This ensures that the systems work as expected before their release

Security and privacy considerations in AI systems

  • AI systems should be secure and respect privacy

  • Machine learning models rely on large volumes of data, which may contain personal details

  • Privacy and security considerations need to be maintained even after training the models and deploying the system

  • As new data is used to make predictions or take action, both the data and decisions made may raise privacy or security concerns


AI systems should empower everyone and engage people. AI should bring benefits to all parts of society, regardless of physical ability, gender, sexual orientation, ethnicity, or other factors.


AI systems should be understandable. Users should be made fully aware of the purpose of the system, how it works, and what limitations may be expected.


People should be accountable for AI systems. Designers and developers of AI-based solutions should work within a framework of governance and organizational principles that ensure the solution meets ethical and legal standards that are clearly defined.