Machine learning applications

Recently, I started working on supporting a state-funded startup that aims to leverage large language models (LLMs) in order to assess qualitative feedback from participants in an organization, with the goal of classifying each response, as well as generating a final report with actionable feedback and tool recommendations. This is a really cool project!

However, the time and scope that we can dedicate to this project is quite bounded—due to our client’s budget constraints. Although we would like to design a nice user journey, and implement a tailored application, we simply don’t have the resources. Fortunately, I found that there exist two Python libraries for rapid development of “machine learning applications”: Streamlit, and Gradio. Overall, both seem capable of supporting a user flow where the application interacts with an LLM, as well as with a custom backend application.

Let us see how this turns out!