Skip to main content

Unlock the full potential of large language models! Dive into the art of Prompt Engineering to harness the power of ChatGPT in boosting productivity across domains by employing various strategies. For instance, ChatGPT as a programming assistant, how to use it in automation activities and speed up daily office’s activities. We will then journey into the secret lives of LLM Agents, overcoming data limitations through the use of external tools and services. This part of the hands-on will be dedicated to a detective-themed puzzle/case designed to perform a simple Extract-Transform-Load (ETL) operation with multi-modal inputs through simple text prompts by connecting the agents to Azure Cognitive Services. The hands-on sessions will cater to both technical and non-technical participants.

Objectives

  • Equip participants with prompt engineering skills for LLM performance enhancement
  • Unlock diverse applications through agents with access to external tools and services

Agenda

09:00 – 09:15 Prompt Engineering Introduction
09:15 – 09:45 Discovering the world of ChatGPT and Prompt Engineering
  • Overview of Large Language Models & ChatGPT
  • A closer look at various Large Language models for Generative AI 
  • The Science of Prompt Engineering
09:45 – 10:30 Hands-on Applications
  • ChatGPT as a Coding Assistant
  • Secretary Mode: Email generation, summarization, and translation
  • Excel Automation with ChatGPT and VBA
  • Data Analysis with ChatGPT for Google Sheets with OpenAI API 
10.30 – 11.00 Break
11.00 – 11.15 Introduction to LLM Agents
  • Understanding the need for using agents with LLMs
  • Definition and components of an agent
  • Connecting to 3rd party tools and APIs
11.15 – 11.30 Hands-on Part 1
  • Simple exercises using tools with LLM Agents 
11.30 – 11.40 Introduction to Langchain & Azure Cognitive Services (ACS)
11.40 – 12.10 Hands-on Part 2
  • Solve a puzzle by performing ETL operations using Agents with Langchain and ACS
12.10 – 12.30 Wrap-up and Q&A

Organizers