What you will learn

  • Learn how to apply advanced prompting techniques such as prompt chaining and ReAct with best practices

  • Build complex LLM applications such as agentic chatbots

  • Learn how to prototype complex LLM applications with no-code tools

Course curriculum

    1. Introduction

    2. Course Objectives and Structure

    1. Flowise AI Installation

    2. Installing Flowise AI

    3. Basic Chatflow with Flowise AI

    4. Exercise

    1. Introduction to Prompt Chaining

    2. Applied Prompt Chaining

    3. Exercise

    1. Food Chatbot with Prompt Chaining

    2. Building the Food Chatbot

    3. Exercise

    1. Introduction to PAL

    2. PAL Demo

    1. Introduction to ReAct Prompting

    2. ReAct Demo

    3. Exercise

    4. ReAct Under the Hood

About this course

  • 21 lessons
  • Projects to apply learnings
  • Earn a Certificate of Completion
  • Intermediate

Instructor(s)

Martin Szummer, Ph.D.

Lead Instructor

Martin Szummer is a course instructor in machine learning with two decades of experience at Google DeepMind, Microsoft Research, MIT, and the University of Cambridge. He has published award-winning research spanning deep learning, kernel methods, and Bayesian methods. At Microsoft, he pioneered algorithms that increased ad revenues by four million dollars in just two months, and at DeepMind, he developed causal machine learning approaches to optimize user engagement long-term. Prior to this, he co-founded and served as CTO of a startup building self-learning voice interfaces, leading product vision, engineering, and fundraising.

More about this course

OVERVIEW

This course focuses on advanced prompt engineering techniques for large language models (LLMs) and how to effectively apply them in various scenarios and use cases. After completing this course, students will have a good grasp of commonly used advanced prompting techniques like prompt chaining, PAL, and ReAct, including how to effectively apply them in their use cases.

PREREQUISITES

If you are new to prompt engineering, we recommend completing the Introduction to Prompt Engineering course (also available to all Pro members). 

The main tool you will use in this course is Flowise AI (no-code tool) with some occasional use of basic Python. You will also need to have a paid OpenAI account for API Keys. More details and instructions are provided in the course.

TOPICS

Throughout the course, students will utilize Flowise AI, to design and apply best practices and common advanced prompting techniques.

Key concepts covered in the course include:

  • Introduction to Flowise AI: Students will be introduced to the popular no-code tool to build advanced chat flows with LLMs
  • Prompt Chaining: Learn how to combine multiple prompt chains in sequence to tackle complex tasks. This module introduces prompt chaining concepts and their practical applications, with hands-on exercises to reinforce your understanding of this advanced prompting technique.
  • Building a Chatbot with LLMs: Apply your advanced prompt engineering skills to create a practical food recommendation chatbot. This hands-on module guides you through, step-by-step, how to build a functional chatbot using prompt chaining techniques and Flowise AI.
  • PAL: Discover PAL (Program-Aided Language Models) and see how it combines LLMs with programmatic reasoning. Through an introduction and live demonstration, learn how this powerful technique enables AI models to solve complex problems by generating and executing code.
  • ReAct Prompting: Explore ReAct (Reasoning and Acting) prompting, a technique that combines reasoning with action planning in LLMs. Through demonstrations and hands-on exercises, learn how to implement this advanced prompting method and understand its inner workings for more effective LLM interactions.
  • Agentic Food Chatbot: In this module, students will apply all the learnings and build an advanced food recommendation chatbot using prompt chaining, an agentic component using ReAct, and Flowise AI. This practical module demonstrates how to build, step-by-step, a more autonomous and intelligent LLM-powered chatbot.