Gathering insights from business documents is very time consuming, and exahusting. That is why we introduce Haystack, to make the journey easier and faster.
First make sure to install the requirements running the code:
pip install -r requirements.txt
After the dependencies are installed, run the command:
fastapi dev app.py
To access the interactive API docs:
http://127.0.0.1:8000/docs
- Query Documents: Users can chat with their own documents to get insights using the powerful Gemini LLM
- Authentication: Users can register and log in through Google's Firebase, which enhance security and scalability
- Document Analysis: Set of machine learning tools, to get more abstract or numerical insights from the documents.
- Keyword counting: Tool that counts all the occurrences of a keyword (and its variants), and records the page location and adjacent text.
- Topic modeling: Tool that helps visualizing the most imporant topics in the documents and their related keywords
- Concept frequency: Tool that helps to visualize the most frequent concepts in the documents.
- Back-end: FastAPI, MongoDB
- Front-end: Flutter
- RAG: Langchain, Gemini LLM and embedding, MongoDB
- Authentication: Firebase
- Hosting: ?