AI Development for Non-Developers: Building a Simple GPT App in Python

Overview

Explore AI development for non-developers in this comprehensive video tutorial. Learn to create a simple Python app using a GPT model, gaining insights into generative AI technologies. Set up your computer, create Azure resources, deploy models, and use the playground. Understand the three key roles in AI interactions and build your own GPT app with authentication. Address memory limitations, implement user input handling, and manage conversation history. Track token usage, test recall capabilities, and review model metrics. Conclude with a code review, summary of steps, and a challenge to further enhance your AI development skills.

Syllabus

- Introduction
- Setup on our computer
- The Azure resources
- Creating a new hub
- Creating a new Azure OpenAI resource
- Adding connections to your hub
- Deploying a model
- Checking model availability
- View model key and endpoint
- Using the playground
- Three roles
- Creating our own GPT app
- How to authenticate to the AI service
- Role on the OpenAI resource
- Lack of memory problem
- Adding memory
- Adding list of dictionaries and system prompt
- Allow repeated user input
- Adding user input to the dictionary list
- Sending the list to the completion response
- Extracting just the message content in response
- Exit message
- Testing with message list
- Do we have full memory and context?
- Add assistant responses to list
- Tracking number of tokens
- Output the history at the end
- Total recall test
- Viewing model metrics
- Code review
- Summary of steps
- Next step challenge
- Use an orchestrator!
- Close