Inspiration DocQuery AI was inspired by the need for a tool that makes finding information in PDFs easier. We wanted to create something that could quickly analyze multiple PDF documents and provide clear and human-friendly answers. We aimed to use AI to simplify how people interact with complex and lengthy PDF, TXT, Word and CSV documents.
What it does DocQuery AI is a sophisticated chatbot designed to analyze multiple documents simultaneously. Users can upload multiple PDF files, and the AI extracts, processes, and embeds the text. It then allows users to ask questions related to the content of these documents, providing precise answers in real time.
How we built it We built DocQuery AI using Python, leveraging Streamlit for the web interface. A key component of our solution was the integration of Google Generative AI for text embeddings, which allowed us to efficiently process and analyze text from multiple PDF documents. Additionally, we utilized FAISS, a vector store, for indexing and retrieval, ensuring fast and efficient document search.
Challenges we ran into One of the biggest challenges was selecting and optimising the correct and most accurate AI model. We also faced difficulties in ensuring the accuracy and speed of the answer-generation process. Additionally, integrating different AI components posed technical hurdles that required careful debugging and testing.
Accomplishments that we're proud of We are proud to have developed a robust AI-driven tool that successfully processes and analyzes multiple PDF documents. Achieving high accuracy in answer generation and making the tool user-friendly were significant milestones.
What we learned Through this project, we deepened our understanding of AI applications in document analysis and natural language processing. We learned how to optimize AI models for specific tasks and integrate multiple technologies seamlessly.
What's next for DocQuery AI Moving forward, we aim to enhance DocQuery AI with more advanced AI capabilities. We plan to integrate more sophisticated question-answering models and improve the scalability of our solution.
Built With
- faiss
- googlegenerativeaiembeddings
- langchain
- python
- streamlit
Log in or sign up for Devpost to join the conversation.