GEMINI_API_2

What it does

GEMINI_API_2 is an innovative application designed to unlock and harness the wealth of information from NASA's Apollo 11 mission archives. Our app processes various types of multimodal data, including text, images, video, and audio files, to provide comprehensive insights and summaries about the mission. By leveraging the powerful capabilities of the Gemini API, GEMINI_API_2 offers users an efficient way to access and retrieve pertinent information from the vast NASA archives.
Here’s how GEMINI_API_2 works:
Data Preparation: We start by collecting data from NASA's public archives related to the Apollo 11 mission, including text documents, images, videos, and audio files.
Data Extraction and Summarization: Using Optical Character Recognition (OCR) with Tesseract, our app extracts text from images. The extracted text is then fed into the Gemini API to generate concise and informative summaries tailored for efficient retrieval.
Embedding Generation: The generated summaries are converted into vector embeddings using Gemini's Text Embedding Model. These embeddings numerically represent the summaries, making them suitable for efficient similarity searches.
Vector Database Creation: We store these embeddings in Chroma DB, a vector database that enables fast and accurate retrieval of relevant documents based on similarity searches.
Querying the RAG System: When a user queries the system, GEMINI_API_2 retrieves the most relevant documents based on their embeddings.