LLM Q/A System for T-Shirt Store

I developed an end-to-end Question and Answer (Q/A) system specifically for a T-shirt selling store. The system uses the Google Palm Large Language Model (LLM) in combination with Hugging Face embeddings to ensure accurate natural language understanding. This setup allows the system to comprehend and respond effectively to a wide range of user queries about T-shirt products.

To handle data efficiently, I established a secure connection between Lang Chain (a tool for building language model applications) and a MySQL database. This connection enables fast and reliable data retrieval and storage, ensuring the system can provide timely and accurate information.

To improve the system's ability to answer diverse user queries, I applied Few-Shot Learning techniques. This method allows the model to learn from a limited amount of data, enhancing its capability to generate precise responses without needing extensive training data.

To make the Q/A system user-friendly, I developed a web interface using Streamlit, a Python-based web application framework. The Streamlit UI allows users to interact with the system seamlessly, inputting questions and receiving instant responses about T-shirt products, such as size, color, availability, and pricing.