Yapper

What it does

Yapper addresses the challenge researchers face in manually reading and extracting data from the overwhelming number of scientific papers published each month. This process is not only time-consuming but also prone to errors, making it difficult to systematically gather and analyze relevant information.
Yapper automates the extraction and organization of data from scientific literature, with an initial focus on plant sciences, compounds, and their roles in human health. Powered by the Gemini API, Yapper allows researchers to effortlessly search, filter, and sort through critical information.
How It Works:
* We scrape abstract data from sources like Nature and PubMed.
* Using Gemini's gemini-1.5-flash-latest model, we extract plant names, compound names, and their roles from the abstracts.
* A self-validation loop, also powered by Gemini, ensures the accuracy of the extracted data, cross-referencing it against the original abstracts.
* Data that passes the validation loop is stored in Firebase Cloud Firestore. Algolia automaticaly indexed new entry in the database,
* The frontend, hosted on Vercel, presents the final, validated data to the user.
Yapper is designed to streamline the research process, reducing the effort required to analyze scientific papers and accelerating the pace of scientific discovery.