DataHour: Building a Simple LLM Application: ChatGPT Summarizer

LLMs and ChatGPT are incredibly powerful tools that will revolutionize how we interact with and consume information. One of the biggest benefits of these models is their ability to understand textual data and extract information from that data. 

In this session, we will work to flex these abilities by building an application that can interpret text and provide a summary in return. Here we will use OpenAI’s API, Python, and Streamlit to build a “summarizer” application. This application will take a PDF, send a prompt to Open AI’s API, and generate a summary in response. We will build the application in Streamlit and work to publicly host it. Overall, this work will provide an introductory entry point into building LLM applications and will give you something that you can share and continue to work on and build. 

Jonathan is a Senior Data Scientist at Nielsen, a Computational Social Scientist at UNC Chapel Hill, and a Data Educator with Correlation One. He started in data science over 10 years ago and has worked in areas of Natural Language Processing, Machine Learning, Inferential and Predictive Statistics, Classification, Clustering, and Recommendation Systems. He has also taught on elements throughout the entire data pipeline and has guided/consulted on 100s of data science projects. He loves this field and feels that there are always new ideas to explore, new skills to develop, and new methods to learn!