The concepts of large language models (LLMs) took the world by storm in November 2022, positioning Artificial Intelligence as one of the most invested-in and promising technology sectors.

This guided project will walk you through the creation of a reasoning and acting (ReAct) agent that harnesses the capability of the most prominent LLM in the world, GPT-4, to automate complex tasks that would normally require human reasoning and input. Ever wanted to know how to use large language models to interact with your business infrastructure or automate customer chat queries? This project is for you.

By the end of this guided, ~1-hr long project, you will have created a GPT Assistant in Node/Typescript, that is able to answer questions on real-time information, such as the stock prices, and also answer questions on given input files. You will also understand the fundamentals of creating assistants that you can use and scale for your own business considerations.

We will walk through the process from the beginning, from setting up your environment and API key, to uploading files and testing the limitations of retrieving relevant information, and creating functions that have reliable logic that can be scaled and changed depending on business need.

This project, of intermediate complexity, is intended for those with some background in programming and application development to fully understand the logic and setup, though even business owners and managerial professionals can benefit from the project as we walk through every step and explain the cost-benefit analysis.

Ready to create your own Assistant? Let's go!