Create $\LaTeX$ and pdf résumés that have been tailored to each individual job posting by an AI LLM model

1. Clone this repo.
2. Provide argument values in env file. Instructions are in the file. Save the file as .env.
3. Open Docker. Install from docker.com if not available.
4. Open terminal and cd into the cloned repo.
5. Start the jupyterlab notebook server -

   docker-compose up --build -d
   

6. Open this URL in your browser to access the notebook - http://127.0.0.1:8888/lab. If you had specififed a different port in the .env file, change it accordingly.
7. Open the notebook - resume_builder.ipynb, and follow the steps within the notebook to create tailored résumés.
8. When done, optionally turn down the notebook server -

   docker-compose down -v
   

1. A raw résumé is provided by the user as a yaml file, which includes all their basic information, experience, projects, and skills. This file will be used as a basis for populating the final résumé corresponding to each job post. An example file raw_resume_example.yaml is included with steps for creating your own.
2. I have created LLM prompts in prompts.py for different résumé related tasks, such as, parsing a job post, rewriting a résumé section, extracting skills, identifying improvements, and writing a summary. The prompts are tested using ChatOpenAI models from Langchain.
3. Based on a job post, the user can perform some or all the above tasks to generate a new yaml file with résumé data that is the most relevant for the job opportunity.
4. The yaml file is then converted to $\LaTeX$ using a jinja2 template. Templates are provided in the templates directory.
5. Finally, the $\LaTeX$ file is converted to pdf using the texlive distribution installed in the Dockerfile.

This github repo is configured to not track any directory starting with my_, so these directories can be used to store personal data without inadvertantly pushing it to github.

Suggestions are welcome. Please create an issue, or fork and PR if you would like to share any improvements in the code or prompts.