From the course: Python for AI Projects: From Data Exploration to Impact

Unlock the full course today

Join today to access over 24,900 courses taught by industry experts.

Improving GenAI performance

Improving GenAI performance - Python Tutorial

From the course: Python for AI Projects: From Data Exploration to Impact

Start my 1-month free trial Buy for my team

Improving GenAI performance

“

- [Instructor] As we continue to evolve our AI system, it's important to step back and talk about the different ways we can structure and optimize workflows in GenAI. Let's begin with two distinct architectural approaches. The first one is a structured AI workflow. This is where you define every step of the pipeline explicitly, from retrieval, to embedding, to prediction, to prompting. Each part of the system follows a predictable path that you can control. It's precise, reliable, and ideal for deterministic flows where you want full visibility and control. The second is a more autonomous or agentic approach where an LLM plays a more active role in deciding what to do next. Instead of simply following steps, the model can choose which tools to call, what actions to take, and how to respond based on the evolving conversation or task. This is where frameworks like LangChain come in, which we've touched upon lightly before.…

Contents