/project-mexitc

Digital Twin for job recruiters to chat with.

Primary LanguageJupyter Notebook

Review Assignment Due Date

image image

Authors:

  • Ruy Guzmán Camacho
  • Fernanda de León
  • Karla Mondragón

Digital-Twin

Welcome to your digital twin. This project is your digital persona, aimed to impress the recruiters and make it easier for them to know about your technical skills, projects and experience.

Instructions

Virtual Environment Setup

Create a virtual environment to avoid problems with package incompatibilities and not interfere with other projects you may have.

python3 -m venv digital-twin  
source digital-twin/bin/activate 
python3 -m pip install -r requirements.txt 
# When you want to leave virtual environment 
deactivate

Environment variables

Create a .env file with the structure presented in example.env

OPENAI_API_KEY = 
PINECONE_API_KEY =
PINECONE_ENVIRONMENT =
GIT_TOKEN =

How to run

This projects relies on 3 main sections

  1. App: contains the main part of the project, UI, modules for our agents and vector database retrievals and logic for the chat workflow.

  2. Notebooks: jupyter notebooks used during the development to test new features in an ordered way. Contains the logic to upload a pdf to a pinecone vector database.

  • notebooks/StudentPrompt.txt : Initial Prompt defining OpenAi behaviour during interaction.
  • notebooks/pinecone.ipynb: Practice with pinecone environment.

  1. Scripts: python scripts used to fetch & process data from APIs and further upload to pinecone

    3.1: Deepgit : Deep analysis to github stats

    • scripts/deepgit/deepGitHub : extracts github commits and code additions from a repository and a user
    • scripts/deepgit/test.py: filters extracted commits to OpenAi to identify code strengths and weaknesses

    3.1.1: codeReview: txt with strengths and weaknesses - scripts/deepgit/codeReview/strong.txt : Strengths obtained from OpenAi Github anaylsis - scripts/deepgit/codeReview/weak.txt: Weaknesses obtained from OpenAi Github analysis

    3.2: Github

    • scripts/github/account_details_Ruy-GC.txt: github main rofile data
    • scripts/github/github_account.py: script to extract gitub main profile data

    3.3: Upload

    • scripts/upload/gh_to_pinecone.py : save github main profile data into pincone

Run chainlit app

cd app/src/
chainlit run app.py -w
#opens the app in the port 8000 by default

Run python scripts

#cd into the folder where the script is 
python3 script_name.py