Programming Generative AI Applications
 
Generative Artificial Intelligence is a class of AI systems that learn patterns from data and use them to create new content across text, images, audio, and video. At the core of most generative AI applications is a large language model (LLM)—a massive neural network, often with billions or trillions of parameters, trained on extensive datasets combined with human feedback.
 
To solve real business problems, these models typically need to be adapted to a specific domain. This customization can be done in two primary ways:
Retrieval-Augmented Generation (RAG): Connecting the LLM to curated business data—such as SOPs, financial documents, reports, or marketing assets—so it can retrieve and use this information during generation.
Fine-tuning: Retraining the model on additional examples so it learns new behaviors, styles, or knowledge tailored to the organization.
 
Building generative AI into business workflows requires gathering relevant data, structuring it properly, and configuring the LLM—either through RAG pipelines or fine-tuning—to ensure it generates accurate, context-aware outputs.
 
This course provides a practical, hands-on introduction to designing and assembling customized generative AI applications. Using OpenAI’s GPT as the base model, you will learn how to implement RAG, perform simple fine-tuning, and integrate these capabilities into real AI systems.
 
Approximately half of the course is dedicated to a group project in which you and your team will build a fully customized AI solution from the ground up. All programming is done in Python using tools such as TensorFlow, LangChain, and FAISS to give you real experience with modern AI development workflows.