Explore what artificial intelligence is and how machines imitate human thinking by using data and algorithms, learning over time to solve problems with a model.

Learn what generative AI is and how it differs from traditional AI and ML. See how GPT models generate new content, with DALL-E and Whisper as examples.

Explore what OpenAI is and why the OpenAI API is needed, and learn how to access models for text, images, and audio via Python or JavaScript libraries.

Explore how ChatGPT uses large language models like GPT-3.5 or GPT-4 to answer prompts, generate content, and assist tasks, while understanding its data cutoff, biases, hallucinations, and ethical considerations.

Learn to set up and use ChatGPT via OpenAI in Azure, compare GPT 3.5 and GPT four, and tailor responses with prompts and custom instructions.

Learn how prompt engineering guides models through carefully designed prompts, using action verbs and explicit context, audience, style, and length to produce precise outputs.

Learn how to use ChatGPT for text summarization, applying prompt engineering to set output limits, structure, and focus, with practical examples from finance and customer service.

Learn to use text expansion to generate detailed content from brief ideas, tailoring tone, length, structure, and audience for emails, reports, and other documents, with practical prompts and examples.

Use chatgpt to research and plan a multi-step travel itinerary, from stuttgart to london via germany and switzerland, with prompt engineering, itinerary updates, and budget estimation.

Learn how ChatGPT acts as a universal programmer, generating, documenting, explaining, and debugging code across languages such as Python, Java, and C#, with practical examples like filtering courses by rating.

Explore how to set context, define a role, and prompt ChatGPT to classify customer tickets into technical issue, billing query, or product feedback and route them to the appropriate teams.

Discover how ChatGPT powers marketing campaigns, from teaser trailers to social buzz and influencer partnerships. Learn prompt engineering and generative AI basics for developers and managers.

Explore how ChatGPT functions as a global translator, translating text between languages like Spanish, English, and Hindi. Learn simple steps to use the translate feature in a chat window.

Learn to apply conditional prompting to generate a title only for English text, and to suppress title output when the input contains the keyword technology.

Learn to control output structure in ChatGPT using conditional prompts, lists, tables, and structured paragraphs with headings and subheadings.

Leverage ChatGPT to generate sample data for testing by producing a csv with ten customers (id, name, age, email, phone) and convert it to json for a restful api.

Explore how Azure OpenAI service provides access to OpenAI models, including GPT-3.5 and Codex, with DALL-E models, Azure security, private networking, RESTful APIs, and a studio for deployment and fine-tuning.

Learn how to request access to the Azure OpenAI service, requiring a company email and an Azure subscription, then select text and code models, Dall-E, and Whisper.

Explore the Azure OpenAI service documentation and model list, review region availability for GPT four and GPT three five, and learn to create resources and apply fine tuning.

Explore pricing quotas and limits for azure openai models, view region-specific USD pricing, and learn GPT three five turbo costs per 1000 tokens at 4K context and quota mechanics.

Deploy an Azure OpenAI resource in the portal, choose East US region and pricing, then access OpenAI Studio to work with text and image generation models.

Learn to deploy, test, and customize OpenAI models in Azure OpenAI Studio, including GPT-3.5 turbo, text embeddings, and DALL-E, using chat, completion, and custom data training.

Deploy a text model in Azure OpenAI Studio by creating a new deployment, selecting GPT three five turbo, setting a unique name, and reviewing quotas before opening the playground.

Fine tune the model with the chat playground by using system messages and examples to shape behavior across the Azure OpenAI assistant setup, chat session, and configuration panels.

Install and use Postman (a RESTful API testing tool) and PyCharm Community Edition to access Azure OpenAI APIs from Postman and Python code, preparing setup for the course.

Learn to access a deployed azure openai model via postman, using http post to the restful endpoint with a json body and api key in headers for chat completions.

Learn to control an Azure OpenAI chat model by adjusting max tokens, temperature, top_p, stop sequences, and penalties. Understand the 4000-token limit shared by system messages, examples, prompts, and responses.

Create a new Python project named 'Azure Open AI demo' in your IDE, configure Python interpreter, and install the official open AI package version 1.3.7 to enable open AI API.

Configure the api key by creating an environment variable named open ai underscore api underscore key on Windows or Mac, and read it in Python with os.getenv.

Learn to access the deployed text model from a Python app with the Azure OpenAI client. Build a chat completion call using system and user messages and parse JSON response.

Learn how to generate multiple responses for a single prompt by setting a system message to bake a cake, requesting three recipes, and looping through choices to print each content.

Deploy a trained model as a web app with Azure Active Directory authentication, configure pricing and resource settings, and test the chatbot in the Chat Playground before sharing access.

Compare the chat playground and completions API, showing how prompts drive task generation like emails, job descriptions, and translations without system or chat messages.

Detect harmful content in prompts and outputs across hate, sexual violence, self-harm, and violence. Customize filters in the studio; the default is medium, and you cannot reduce severity.

Discover how to create a customized content filter, name it, set severity for hate, sexual self-harm, and violence, and apply it to a model deployment to block problematic content.

Learn to use text embeddings to convert words and sentences into vectors that capture meaning and relationships. Compare embeddings with cosine similarity for tasks like document similarity and search.

Deploy an embeddings model using Azure OpenAI Studio, creating a new embedding deployment named embed_demo and preparing to fetch text and sentence embeddings with Python.

Install NumPy to access the dot method for computing cosine similarity between embedding vectors, enabling comparison of word and sentence embeddings in your Python project.

Learn to retrieve word and sentence embeddings using the deployed Azure OpenAI Studio embed_demo model by calling the embeddings create method and reading the resulting embedding data.

Learn to compute similarity between words and sentences using embeddings, numpy's dot and cosine similarity, with practical Python demos comparing hot, cold, good, bad, and you are good vs awesome.

Deploy a new Azure OpenAI resource in West Europe to access whisper models, then open Azure OpenAI Studio to use whisper, GPT, and text embedding.

Learn to transcribe audio with a deployed whisper model via the Azure OpenAI API in Python, including environment setup, file handling, and using transcriptions.create to obtain text.

Translate a sample French audio file to English using the translate script and audio translations, turning a customer interaction recording into English text for call center learning.

Learn how to mix audio and text models by transcribing an AWS Lambda mp3 file with whisper, then summarize the resulting text using a deployed text model.

Azure OpenAI does not retrain on data or include it in prompts; it ingests and stores data, builds a cognitive search index, and retrieves via hybrid, vector, or keyword search.

Explore how ingestion processes uploaded data from blob storage, a cognitive search index, or local upload into 1024-token chunks, with optional embeddings for vector search, building a searchable Azure index.

Use a practical use case by loading five data scientist resumes into a chat bot with your own data, then test queries in the playground to help HR filter candidates.

Set up a storage account with a unique name and container in East US, plus a cognitive search service in the same resource group, then delete service to avoid charges.

Learn how to set up hybrid search by enabling embedding models, creating a vector search index, uploading data, and using embeddings to power queries.

Build a smart home hub chatbot by deploying a text model with cognitive search, using the PDFs to answer compatibility, voice command, and mobile app questions.

Delete the cognitive search resource and storage container in your resource group after experimentation to avoid costs, confirming each deletion step and naming the resources as prompted.

Master the fine tuning use case and json message format using system, user, and assistant roles. Prepare training and validation data in jsonl files to tune and validate the model.

Deploy the Azure OpenAI service and customize the GPT 3.5 turbo model by selecting a base model and creating a custom model in the Azure OpenAI Studio, with region options.

Upload training and validation data in Azure OpenAI studio, then fine tune a base GPT three five turbo model into a custom model, monitor training events, and deploy when complete.

Learn to fine tune a model with the OpenAI API using Python: upload training and validation jsonl files, start a fine tuning job on Azure OpenAI, and monitor status.

Fine-tune an OpenAI model via API by preparing data in OpenAI Studio, applying a delay for updates, and deploying the custom model to a playground.

Delete your fine-tuned deployments and models after use to avoid ongoing charges; go to deployments, delete the deployment first, then delete the model to start the deletion process.

Explore image generation and editing with dall-e in azure openai, crafting precise prompts, adjusting size and variations, and applying create and edit methods for hands-on image creation.

Deploy the Dall-E three model with Azure OpenAI Studio by creating a Sweden Central resource, then deploy image_underscore_demo for use in Studio and Python.

Explore generating images from prompts with a deployed DALL-E model in the studio, using the playground, saving prompts, downloading images, and building Python code to invoke generation.

Generate images from Azure OpenAI DALL-E three in Python by calling client.images.generate with the image_demo model and retrieving the image url from the json response.