LangChain Documentation Helper Chatbot - Helper Bot

A Python-based documentation helper that answers marketing/sales questions

Features

• File Ingestion to VectorStore: Ingestion of files into a any vectorstore (Pinecone for now) for search and retrieval capabilities within your documentation.
• Natural Language Processing: LangChain facilitates the creation of complex workflows by chaining together different LLM related operations
• Customizable Templates: Usage of custom templates to pass as prompts to LLM.

Installation

• Clone the repository
• Manually add .env file to the project (running the project without the API keys is impossible)

Three ways to work with the project (to be updated constantly...)

Using Global Python Installation

• Ensure Python 3.11 or later is installed on your system.
• Make sure you have pip installed on your computer
• To install the needed packages: pip install -r requirements.txt
• To run the main file of the program: streamlit run main.py

Using a Virtual Environment (pyenv)

• Make sure to correctly install pyenv according to this article (Ubuntu >= 22.04 preffered): https://medium.com/@aashari/easy-to-follow-guide-of-how-to-install-pyenv-on-ubuntu-a3730af8d7f0
• After installing pyenv, install any desired version of python: pyenv install 3.11.1
• Make sure you have pip installed on your computer
• Create the virtual environmen: pyenv virtualenv 3.12.1 myenv
• Activate environment: pyenv activate myenv
• To install the needed packages: pip install -r requirements.txt
• To run the main file of the program: streamlit run main.py
• To deactivate environment: pyenv deactivate

Using a Docker Image

• There is a Dockerfile inside the repo
• To build a local image: docker build -t image_name .
• To run the image: docker run -p 8501:8501 image_name