📄 CVApp AI

This is a foundational component of a larger AI-powered application that assists in generating customized CVs based on personal experience and job descriptions.
Currently, this part focuses on:

• Securely reading content from a private Google Doc
• Scraping job descriptions from external job boards
• Generating personalized CVs using LangChain + Google Gemini
• Saving the result as a Markdown file named after the job title

This is still in development

---

✅ What I’ve Learned So Far

• How to use the Google Docs API with Python
• How to authenticate using OAuth (with refresh tokens)
• How to manage secrets securely using .env files
• How to structure a clean project with good practices (like .gitignore and token caching)
• How to scrape websites using requests and BeautifulSoup
• How to extract relevant job offer content and save it for later use
• How to build prompt chains using LangChain and Gemini

---

🚀 Getting Started

1. Clone the Repository

git clone https://github.com/olagraczyk/cvapp_ai.git
cd cvapp_ai

2. Google Cloud Setup

1. Go to Google Cloud Console
2. Create a new project
3. Enable the Google Docs API
4. Go to APIs & Services > Credentials
5. Create OAuth client ID credentials (select Desktop App)
6. Download the client_secret.json file

3. Project Folder Structure

Place your credentials as follows:

cvapp_ai/
├── auth/
│   └── client_secret.json      ← Put it here
├── .env                        ← Your environment variables
├── main_app.ipynb              ← Core app logic
├── web_scraper.py              ← Job board scraper
...

In .env file paste

GOOGLE_DOC_ID=your_google_doc_id_here
GOOGLE_API_KEY=your_gemini_api_here

Google Doc Id is placed in the url of the doc. You can find which one it is easly on the Internet

Update the .gitignore file if you don't want your secrets and tokens to be tracked by Git.

.env
auth/client_secret.json
auth/token.json

4. Install Dependencies

In terminal

pip install -r requirements.txt

---

Web Scraping Job Descriptions

You can run web_scraper.py to extract the main content from a job posting page.

Currently:

• The script takes a hardcoded URL (will be updated for dynamic input later in a Streamlit App).
• It saves the job title and main content into job_description.txt.

In the Jupyter notebook prototype, you can enter a job link as input and run the scraping function. This will update the job description for the CV logic later.

CV Generation with Gemini

The app uses LangChain + Gemini to generate a two-part prompt chain:

1. Process job description
2. Generate a full, job-specific CV in Markdown, tailored to the job description.

🛠 To Be Continued...

---

Author: Ola Graczyk