Scholar Assist

Google Scholar Assist is an AI-powered research assistant that streamlines the research process. It leverages the power of Google Scholar to intelligently analyze articles, extract key information, and provide insights to help researchers focus on what truly matters, by improving Search Accuracy and Information Retrieval with Artificial Intelligence.

For this project, I was the lead developer and created the idea to pursue. I was responsible for developing the AI functionalities, and designing the interfaces for the related components. I built an AI model with Vertex (Gemini), designed to enhance the efficacy of Google Scholar by rating the relevance and quality of the papers’ results. It analyzes the content of the articles that are provided to it, to determine their relevance based on user research needs. In the next iterations, I built tools which provide detailed summaries of articles, highlighting key findings and citations, as well as a tool to get specific quotes from the research paper based on your paragraph or research paper content.

This project was a really informative one for my own learning process, as I learned how to further fine-tune AI’s, and I got to expand and employ previous knowledge on component creation from my previous internship, into this project. I also learned more about using Visual Studio Code as a platform, and gained more knowledge and practice with the interface during the process. This project also required a lot of background research into the functionality of Google Scholar, how consumers interact with it, and what improvements it could need, to even find the idea to pursue, so I learned a lot more about Google services during this process as well.

You can learn more and read the code on my git repository.