Analyzing 400 Obama Speeches with GPT 3.5 Embeddings

Thomas Ragger
Published in
2 min readFeb 27, 2023


Machine learning is an exciting field that is rapidly evolving, and it is changing the way we approach and analyze data. At wild, we are constantly exploring new ways to use machine learning to help businesses and individuals extract meaningful insights from their data.

Recently, we utilized the GPT 3.5 embeddings API to parse 400 transcripts of Obama’s public speeches and make them easily searchable through natural language. By using this advanced technology, users can search through hundreds of speeches from one of the most influential leaders of our time using everyday language, which makes analyzing and interpreting vast amounts of data much more straightforward.

Check out the prototype:

There are an increasing number of exciting frameworks & tools being released every day, such as Langchain, GPTIndex, and Pinecone, which are making tasks like data analysis and search easier and more efficient. These powerful tools can help developers build innovative solutions and enable users to understand complex patterns that were previously impossible to uncover.

At wild, we are passionate about helping our clients explore the world of machine learning, build proof of concepts, and realize their ideas. So if you’re interested in learning more, don’t hesitate to get in touch and shoot an email to