Building a Custom OpenAI Chatbot

Via Udacity

Udacity based on 0 ratings

Likes

Overview

This course teaches learners how to build a custom chatbot using Open AIâs GPT models. Topics include API integration, prompt structuring, memory simulation, and context management. Learners will explore chat history persistence, response control, and deployment on platforms like web apps or Slack. Emphasis is placed on user experience, security, and ethical AI usage. By the end, participants can create and deploy interactive, task-specific AI chatbots tailored to business or personal use cases.

Syllabus

Welcome

In this lesson, you'll get an overview of our journey. You'll learn a bit about OpenAI's history and how we can use it. You'll also get a brief introduction to prompt engineering.

Preparing a Dataset with Embeddings

In this lesson, we'll prepare our dataset to make it accessible for machine learning. We'll find the data, load it, clean it up, and use embeddings to create numeric representations of our data.

Finding Relevant Data

In this lesson, we will discuss the differences between keyword and semantic searches. Then, we'll find the relevant data for our queries using cosine similarity.

Querying a Completion Model with a Custom Text Prompt

In this lesson, we'll get to the magic of AI! We'll write a custom prompt that includes the most relevant parts of our dataset and use that prompt to get a customized response.