E-Learning Q/A System

I led the development of a Real-World Q/A System for an e-learning company, beginning with a comprehensive requirements analysis to tailor the solution to their specific needs. The project involved integrating the Google Maker suite for advanced Natural Language Processing (NLP), including setting up API keys and authentication to ensure seamless interaction.

To enhance contextual understanding, I utilized and fine-tuned Hugging Face’s pre-trained models for instructor embeddings. I also established a vector database using FAISS, optimizing it for both speed and memory efficiency, to facilitate the rapid retrieval of relevant information.

The core of the system was a Lang Chain Retriever QA module, which I developed to integrate the vector database and embeddings, ensuring accurate and prompt responses to user queries. Additionally, I created an intuitive Streamlit user interface, allowing users to easily input questions and receive precise answers, thereby improving the overall user experience in the e-learning environment.